TRANING MANUAL ON ULTRASONIC TESTING

COURSE NO. : SA-M-QC-3.6

Prepared by

V.K. Jain, SO/D

Reviewed by

Prashant Puri, STO

A.K. Singh, Head QA

C.M. Mishra, ENC (MT)

Approved by

N. Nagaich Training Superintendent, RAPS 1 to 4

Nuclear Training Centre

Revision (0) Aug., 2002

Next Revision due : Aug., 2007

PREFACE

This training manual on the subject of Ultrasonic Testing is complied and prepared to be used for the theoretical and practical training on the subject. This manual contained self explanatory chapters including the basic knowledge of sound waves. Source of material compiled in the manual taken from the U.T. lecture notes ISNT chapter Rawatbhata Kota. with the little changes as per the requirement of training. I express my sincere thanks to Shri Prashant Puri, STO for giving his valuable suggestion during preparation and Shri C.M. Mishra, ENC(MT) & Shri A.K. Singh, Section Head, QA for their guidance and keep me cheerful to complete the task. I will be failing in my duty if I do not record my thanks to Shri N. Nagaich, Training Superintendent, RAPS 1-4 for his kind guidance, constant encouragement & motivation to prepare this manual.

V.K. Jain
SO/D, NTC.

CONTENTS
Chapter No. 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 UT in A View. Introduction Through Transmission & Pulse Echo Techniques Sound Waves and Wavemodes The Ultrasonic Flaw Detector Equipment Controls Equipment Characteristics Calibration Blocks Thickness Measurement Generation and Reception of Ultrasound Normal and Angle Beam Probes Dual Probes Pulser and Receiver Profile of Ultrasonic Beam Wave Propagation in Material Intraction of Ultrasonic with Material Selection of Test Parameters Accessories Defects Enountered in Materials Tables I, II, III & IV Diagrams Description Page No. 1-3 4 5-7 8-9 10-15 16-18 19-22 23-25 26-27 28-32 33-34 35 36 37-39 40-42 43-46 47-48 49-52 53-58 59-85

Chapter - 1

UT IN A VIEW
1.1 INTRODUCTION Ultrasonic testing (UT) makes use of sound waves and is based on the principle of echo. Sound waves are sent into the material being tested and they are reflected back by the defects in the material. The reflected waves give information about the defect. Sound waves are caused by mechanical vibrations or mechanical oscillations. One vibration or oscillation is also called a cycle. The number of cycles per second (CPS or C/ S) is called the frequency of the sound waves. One cycle per second is called a Hertz (written as Hz). One thousand Hertz are called a kilo Hertz (KHz) and ten thousand KHz (or one million Hz) are called a mega Hertz (MHz). Sound waves with frequencies from about 20 Hz to about 20 KHz are audible i.e. they can be heard by human ear. Inaudible sound waves above 20 KHz are called ultrasonic waves. In UT, we make use of ultrasonic waves. Frequencies from 0.5 MHz to 5 MHz are most commonly used in UT. In special cases frequencies as low as 0.2 MHz to as high as 20 MHz may be used. Ultrasonic waves are generated by the transducer (also called the probe or the search unit) when it is connected to the ultrasonic flaw detector. The waves are generated not continuously but in small bursts or pulses. The pulses are sent into the test object by placing the transducer in contact with the test object. They travel through the material and are reflected back by the far surface (back surface) as well as by any intermediate reflectors such as defects. The reflected waves (or the echoes) are received by the same transducer. The transducer converts the sound waves back into electrical voltages. These voltages are displayed on a cathode ray tube (CRT) of the ultrasonic flaw detector. The horizontal distance on the CRT is proportional to the distance of the reflecting point from the transducer. The vertical height of the reflected pulse or echo is a measure of the energy reflected by the defect and the size of the defect is judged from this echo height. The shape and behaviour of the echo give some information about the type and shape of the defect. A large number of parameters are involved in UT, such as (a) the probe characteristics, (b) the equipment characteristics and control settings, (c) the acoustic behaviour of the test material and the defect.

To make sure that proper parameters have been selected, the procedure is proved on reference blocks or the reference standards which are made of material identical to the one being tested and which contain artificial defects. In addition to flaw detection, ultrasonic testing is also very commonly used for measuring thickness, since the distance of the back echo on the CRT is proportional to the distance of the back surface. A flaw detector can be used as a thickness tester as well. Equipment specially designed for thickness testing alone are also available. However for proper application of UT, we should have a knowledge of a wide range of phenomena dealing with the physics of ultrasound and the interaction of ultrasound with materials and defects. For example : The ultrasonic waves diverge as they pass through a medium and the divergence depends on the ultrasonic frequency and the size of the transducer. Due to this divergence the intensity of the ultrasonic beam decreases rapidly with distance. The intensity also gets reduced due to absorption and scatter : this depends on the frequency as well as the metallurgical structure of the material and its crystallographic characteristics. The waves get reflected and refracted at interfaces; the partition of energy between the reflected and refracted components depend both on the acoustic characteristics of the two media as well as the angle of incidence; The defect may reflect or scatter the incident waves depending upon its size and shape as well as the wavelength of the incident waves. Applications may demand different degrees of sensitivity and resolution and different thicknesses of materials may have to be inspected. Sometimes these requirements may be contradictory and a compromise must be arrived at, sometimes meaningful UT may not be possible. The operator has many parameters to choose, such as the transducer material, its size, the frequency, damping, amplifier characteristics, noise suppression, single element or twin element transducers, the beam angle, the beam shape, the direction in which the beam is sent into the test object and the parameters of the echoes to be measured.

And practical skill is as important as theoretical knowledge. The operator should acquire a skill for picking up and maximising indications, develop a 'feel' about the echoes, should be able to repeat the results and learn to extract maximum data from the echo. This skill can be acquired only through sufficient practice and experience. While further lectures will help us to acquire theoretical knowledge in a systematic way, the practical will initiate us into the proper setting up of the equipment for a given application and to scan a given material using a given procedure and report the results.

Chapter-2

THROUGH TRANSMISSION AND PULSE ECHO TECHNIQUES

As explained in the earlier chapter, UT is based on the principle of echo. But, the first ever UT system suggested (60 years ago) was similar to radiography. In this system a continuous beam of ultrasonic energy is sent into the material. Any defect will reflect the incident ultrasonic waves and thus cause reduction in the transmitted energy. This technique is now called the 'through transmission technique' or simply the 'transmission technique'. (Fig. 1) During the second world war 'radar' was used for detecting or locating airplanes. In these systems pulses of radio waves were sent into space and the reflected pulses were received and analysed. This idea of using pulses was adopted in UT and the 'pulse-echo technique' was born. The first ever commercial UT flaw detector, made in 1942, was based on this principle. (Fig. 2) The advantages of pulse echo technique over through transmission technique are : 1. Time taken for the pulses to return can be measured and hence the distance of the reflector (i.e. depth at which the defect is located) can be measured. Access to one side of the object is sufficient. The transmitter can act as the receiver also since it is not continuously transmitting; thus one probe is sufficient. The problem of aligning two probes precisely is eliminated.

2. 3.

4.

Nowadays the pulse echo technique is almost exclusively used. Through transmission technique ( using pulses) is used mainly in automated inspection of thin components and in some other special cases. To transmit the vibrations of the probe into the material, the probe is kept in contact with the material. If there is even a very thin air gap between the probe and the test material, all the vibrations will be reflected back into the probe. To avoid this and thus to improve the transmission of the waves into the material, a thin layer of a liquid (oil, water and glycerine) or a semisolid (grease, gel) is applied on the test object. This material is called the 'couplant'. For the same purpose ( of transmitting ultrasonic energy into the material being tested), a column of water can be interposed between the probe and the material, or the material can be immersed in water. In these cases the water acts as the couplant. We shall learn more about this later.

Chapter-3

SOUND WAVES AND WAVE MODES

We said that acoustic waves are produced and sustained by mechanical vibrations. When an acoustic wave travels through the medium, the particles in the medium vibrate. A vibration is a periodic phenomenon. When a particle vibrates, its displacement (movement from the normal position) increases and decreases periodically. A wave is characterised by certain quantities :The AMPLITUDE (A) is the maximum displacement of a particle. See Fig.3 The WAVE LENGTH () pronounced ('Lambda') is the distance between two successive similar positions in the wave. See Fig. 3 The PERIOD (T) is the time taken by the wave to travel a distance equal to a wave length. The FREQUENCY (f) is the number of waves travelling acrosses a point in one second. It is also the same as the number of vibrations executed by the particles per second. The following equations connect the different parameters : V= f F=

1 T

There are different types of ultrasonic waves ; these are called different waves MODES. In LONGITUDINAL WAVES or COMPRESSIONAL WAVES, the particles vibrate parallel to the direction of propagation of the waves. See Fig. 3-A In TRANSVERSE WAVES or SHEAR WAVES the particles vibrate perpendicular to the direction of propagation of the waves. See Fig. 3-B RAYLEIGH WAVE or SURFACE WAVES travel only on the surface and they exist only upto a depth of one wavelength in the material. The particles move in elliptical paths. See Fig. 3-C LAMB WAVES or PLATE WAVES travel in thin plates where the thickness is smaller than the wavelength. The motion of particles in this case is very complex. Of these four modes, the first two are most commonly used in UT. Surface waves are sometimes used and Lamb waves rarely.
5

Each of these modes have different velocities. These velocities depend upon the physical density and the elastic constants of the material. The velocity relationships are given below :

V0 =

E P

VT =

E 1 P 2(1 + ) =

G P

Where, V0 VL VT VR Y G = = = = = = = = Velocity of longitudinal waves in a thin rod Velocity of longitudinal waves Velocity of transverse (or shear) waves
L

0 .87 + 112 . Velocity of surface (Rayleigh) waves (1 , ) R =

Young's modulus of elasticity Shear modulus of elasticity Poisson's ratio Physical density

1 ++ ) (1 2 ) (1
T

From the above equation, it is seen that velocity depends only on the characteristics of the medium in which the sound waves travel and not on other parameters like frequency, source of sound etc. For a given material, longitudinal waves have the greatest velocity is smaller than longitudinal wave velocity and it varies from 0.4 to 0.7 of the longitudinal wave velocity. Surface waves velocity is smaller than shear wave velocity and is approximately 0.9 times shear wave velocity. As for lamb waves, there are different kinds of the each with different velocity. The relationships are complicated and are usually represented graphically, Since they are not of much use to us, we shall not discuss them here.

AMPLITUDE, PRESSURE AND INTENSITY : The particle motion creates periodic compression and refraction in the material. Thus the wave can also be considered as a pressure wave. The relation between pressure (p) and particle amplitude (A) is given by : P zA where, z is called the 'acoustic impedance', and z= density x velocity The intensity (I) of the waves which is the energy carried by the waves per second across unit area is given by I zA2 > 10 watt/cm2

Chapter-4

THE ULTRASONIC FLAW DETECTOR (Block Diagram)

We learnt that the most common technique employed in UT is the pulse echo technique. The first ever commercial flaw detector, made in 1942, was based on this principle. Great improvements have been effected since then in the circuitry to make the system more versatile and more capable. However, the basic construction of the ultrasonic flaw detector has not practically changed since the first machine. This chapter describes how a standard flaw detector works. The Cathode Ray Tube : Now, since a cathode ray tube (CRT) forms an essential part of the flaw detector, we should first understand the functioning of the CRT. A CRT is essentially and evacuated glass tube. The cathode filament 'C' is heated electrically and it emits electrons. (Such emission of electrons due to heating is called thermionic emission ). The emerging electrons are focussed into a thin beam by the focussing coils 'F' and the electrons are accelerated by a high voltage applied to the anode 'A'. The accelerated electron beam finally strikes the flat circular and 'S' of the CRT. This flat end of the CRT is called the CRT screen is coated with a fluorescent material. The area where the electron beam strikes the screen will be seen as a luminous spot. 'HH' and 'VV' are two sets of electrodes. If a voltage is applied to HH plates the electron beam will be deflected horizontally and if a voltage is applied to the VV plates, the electron beam will be deflected vertically. See Fig. 4 In a flaw detector, the horizontal deflection is used to measure time. Hence, a voltage varying linearly with time is applied to these plates. Such a voltage is called a 'sweep voltage' , a 'ramp voltage' or a 'saw-tooth voltage'. See Fig. 4-A When one cycle of sweep voltage is applied, the electron spot will travel at UNIFORM speed from left to right and instantaneously reach back to its starting point. Since the speed is uniform, the distance of the spot along x-axis is proportional to time. If this cycle is repeated many number of times a second, we shall see a line on the CRT due to persistence of vision. Simultaneous with a sweep voltage, if a transient voltage (i.e. a voltage of short duration) appears on the VV plates, the trace on the CRT will also instantaneously deflect up and down. This vertical deflection on the CRT is called a 'pip'. In UT, such pips are also called echoes, reflections, vertical deflections, reflected pulses etc.

The Flaw Detector - Block Diagram : Having understood how a CRT functions, let us now study how a flaw detector works. We have a pulser 'P' which gives out a high voltage electric pulse of short duration. The electric pulse excites the transducer 'T' and the transducer generates a short pulse of ultrasonic waves. The waves travel into the material under test and get reflected by a defect or the back surface. The reflected waves reach back to transducer, and get converted into an electric voltage. This voltage is amplified by amplifier 'A' and fed to the vertical deflection plates VV of the CRT. This will cause a 'pip' on the CRT as explained earlier. See Fig. 4-B When the pulser 'P' emits a pulse, the sweep circuit 'S' is simultaneously activated and this sweep voltage is connected to the horizontal deflection plates HH of the CRT. Hence the electron beam starts tracing the horizontal line on the CRT. The pip appears after some distance on the x-axis since there is an interval of time between the start of the sweep and the return of the ultrasonic pulse; and this distance will be proportional to the distance of the reflector from the front surface of the test-object. A part of the pulse emitted by the pulser also goes through the amplifier and therefore we get a fairly large vertical deflection 'TP' at the start of the horizontal line on the CRT. This vertical deflection 'TP' is called the initial pulse, the transmitted pulse or the main bang. Since the pulses and sweep generate voltages periodically, they must be made to act in unison; i.e. the time difference between the start of high voltage pulse and the start of the sweep should be zero or a constant. In other words the pulser and the sweep circuits should be synchronized. This is done by the synchronizer (or clock or timer) 'C', which gives out an electrical voltage periodically which activate the pulser and sweep circuit simultaneously or with a constant time delay. This, in principle, is how a UFD works. We have discussed only the essential components of the ultrasonic flaw detector. There are many components like the rectifier, clipping, sweep delay, etc. about which we shall learn later.

Chapter-5

EQUIPMENT CONTROLS
In the last chapter we saw how the Ultrasonic Flaw Detector functions, by using the block diagram approach. Now let us learn what are the different equipment controls (knobs) which the UT operator will be handling in his day-to-day work. The Sweep Controls : The sweep voltage deflects the electron beam horizontally. Hence all the sweep controls will affect the CRT screen picture only in the horizontal direction. Range Control : We saw that the sweep is a voltage which varies linearly with time. If the sweep voltage rises to the maximum in a short time, the horizontal axis on the screen will also represent a short duration of time; and in terms of distance travelled by the ultrasonic beam, it will represent a short distance of travel. See Fig. 5 The sweep control alters the rise time of the sweep voltage and thus varies the distance or range of thickness represented in the CRT. Therefore this control is called the 'range control'. Since steel is the most common engineering material, this 'range' is usually directly calibrated in mm of steel. In some equipment, there can be two controls for range, one of fixed steps and the other continuously variable. In such cases the control of fixed steps is calibrated in mm of steel and the continuous control can be used for setting accurately the same mm range for different material. Thus the continuous control is sometimes known as materials control. Since the materials are characterised by velocity of sound for this control, it is also known as the velocity control. Whatever the name given and whether the knob is a single or double one, one should remember that this knob is (or these knobs are) used to set the horizontal scale of the CRT to represent a given thickness ( or a given metal path) of a given material by varying the sweep time. The range is always set by using known thicknesses of the material and the numbers indicated by these controls or dials are not to be taken for granted. Delay control : We said that the pulser and sweep are made to start simultaneously by the clock circuit. However, when the range control is used (i.e. when the sweep rate is varied) a time difference gets introduced between the start of the sweep and the start of the pulse. This is physically seen on the CRT by the shift of the initial pulse to the left or to the right on the horizontal scale of the CRT. To off-set this error a control is available which can change the delay between the pulser and the sweep circuit. This control is called the delay control, zero shift control or the horizontal position control.

10

Also, one sometimes wishes to ignore the first few mm or cm of the material or would like to observe, say, only between the first and the second back echoes, or sometimes, a column of water or a block of solid material is interposed between the probe and the material under test. In such cases no information is obtained till the time the ultrasonic beam reaches the surface of the test object. Here also one would like to omit the initial part of the CRT display. The delay control is used for such cases also. See Fig. 5-A It should be noted that the delay control shifts the whole trace bodily to the left or right. Since the sweep rate is not changed, the horizontal distances between any two pips remain unaltered. On the other hand, if the 'range control', or sweep rate, is altered the distance between pulses is increased or decreases. Interface trigger control : Sometimes the test object is immersed in water. Since zero time is counted from the time the waves enter the material, the echo from the front surface of the object is to be taken as zero point on the CRT. This can be done by zero shifting (or delaying) the sweep as explained above. However, if the water column distance varies, the zero setting will be affected and the delay also has to be varied. The interface trigger control triggers the sweep only when the interface arrives back. Thus it keeps track of the varying or 'floating' zero point and always adjusts the first echo (i.e. the echo from the top of the test object) to be at the zero of the CRT. This control is also known as the echo-follower control or the echostart control. This control is not commonly provided in all machines and one should ask for it if needed. The Pulser Controls The pulser controls alter the height and shape of the pulses. Pulse energy control : Ultrasonic pulses are generated by giving short duration electric pulses to the transducer. If the voltage is higher, the ultrasonic vibrations also will have higher amplitude and hence the energy content of the ultrasonic pulse will also be high. Pulse energy control increases or decreases the voltage applied to the transducer. The increase in the pulse energy due to the application of the higher voltage, increases not only the amplitude of vibrations but also the length of the ultrasonic pulse so that the pulses become broader. See Fig. 5-B Damping Control : If an external force is applied to a freely vibrating system, the vibrations stop within a shorter time. This loading which inhibits the vibrations is called 'damping'. Therefore, damping will make the pulses narrower. Simultaneously the height of the pulses or, in other words, the energy of the pulses will also decrease. See Fig. 5-C Damping can be mechanical or electronic. Mechanical damping is achieved by adding a heavy load to the back of the transducer. This is in-built into the transducer and
11

hence cannot be altered. Electronic damping is variable and is achieved through the damping control. It is also to be noted that damping alters the frequency of vibration of the transducer. It makes the transducer vibrate simultaneously over a wide range of (or over a 'broad-band' of) frequencies. When the transducer is damped it becomes a broad-band transducer. Higher the damping, broader is the frequency band and shorter the pulse length. The Amplifier Controls The amplifier controls, like the pulser controls, affect the shape and height of the pulses. Gain Control : The amplifier amplifies the received signal from the transducer. Since the amplitudes of the received signals can vary over a large range (ratios of 10,000 are not uncommon) and any signal of interest should be represented on the same small CRT screen, we need to control the amplifier gain. The gain control can be of two types; the uncalibrated and the calibrated. The uncalibrated control does not have any meaningful calibration of the gain level and hence is not useful calculations or measurement. This is generally used to set the height of echoes to precise levels. This control should be locked and should not be used once the testing has commenced. The calibrated gain control, on the other hand, is calibrated in numbers representing the ratio of amplification. Since the ratios can vary a large range (10,000 as mentioned earlier) and since we are interested in relative ratios rather than the absolute ratios with respect to some fixed level, it is more convenient if the logarithms of the ratios are used. This logarithmic ratio is specified in terms of decibels and the linear ratio is given by the formula : dB=20 log10 (A1/A2) where A1/A2 is the linear ratio of amplitudes. In certain equipment we have the ' attenuation control' instead of the gain control. Attenuation (by the amplifier) is the reverse of amplification. Increase in gain means increase in amplification whereas increase in attenuation means decrease in amplification and subsequently decrease in the pulse heights as seen on the CRT. Note : The attenuation by the amplifier should not be confused with attenuation of ultrasonic energy ( i.e. reduction in ultrasonic energy as it propagates in the material). Wave form Display Control : Usually the pulses are rectified before they are amplified and presented on the CRT. Full-wave rectification is usually done. Some
12

equipments give the rectifications and full-wave rectification. See Fig. 6 It is also sometimes useful to have the unrectified pulse on the CRT. Some equipment give this option also. The unrectified wave form is called the RF (radio frequency) form. The wave form presentation control helps the operator to choose the presentation of the required wave form on the CRT. Suppression Control : Sometimes one gets small, clustered and constantly varying echoes on the CRT which are not due to defects. Such echoes are called 'grass' echoes or 'noise' echoes or 'hash' echoes. These echoes cause fatigue to the operator since he has to be looking for defect echoes and not these noise echoes. See Fig. 6-A The suppression control (or reject control or clipping control ) cuts off a fixed amount of height (or voltage) from all the signals so that the noise is removed from the CRT. It must be noted that this control affects only a cosmetic change on CRT and the following warnings should be remembered while using this control : If suppression is applied before the testing starts : 1. 2. 3. Information regarding the scattering nature of the material will be lost. Small signals also will get cut-off along with noise. The echo ratios get affected by this control and hence suppression should not be used when linear ratios of echoes are to be measured from the CRT.

Of late, equipment have come into the market which are having linear suppression control. This linear suppression cuts of echoes which are smaller than the predetermined level and does not affect the heights of the remaining echoes. Therefore the third warning above is not applicable in the case of linear suppression. Filtering control : If the ultrasonic pulses sent into material has high frequency components, the reflected pulses seen on CRT can have a serrated look. This shape can be inconvenient while reading thickness (or position of the echo along the x-axis of the CRT) and one would like to have a smoother looking pulse. This is achieved by filtering high frequency components. See Fig. 6-B This control also should be used with caution because the serrated nature of the echo may be due to the irregular shape of the defect and this information is lost by introducing filtering.

13

Frequency or bandwidth selector control : Though we talk of a probe as if it generates waves of a single frequency, actually a band of frequencies are generated by it. The band can be narrow or broad depending upon the probe and equipment settings (refer to 'damping control' above). Now the amplifier can be capable of receiving a wide range of frequencies with equal efficiency; in this case the amplifier is called a broad-band amplifier. The amplifier which does not have this capability has to be adjusted every time to receive the band of frequencies which has been transmitted by the probe, such amplifiers are called narrow-band amplifiers or tuned amplifiers. Equipment with tuned amplifiers will have a frequency selector control. This controls alters only the amplifier and should be set so as to receive the transmitted frequencies properly. This does not affect the frequency of the transmitted waves, which is determined by the probe and the damping control described earlier. Some equipment give the options of both broad-band and narrow-band tuning. Time corrected gain (TCG) : The intensity of the ultrasonics beam decreases with distance due to many reasons. Therefore, echoes from the defects of same size will also decrease with the distance at which they are situated. Since defect size is judged from the echo height it will be useful to make the height of echoes from same size defect equal. The electronic time corrected gain of the amplifier is increased with time (like the seep voltage) to compensate for the loss of energy with distance. Since the reduction of echo height is not linear with time a simple linear TCG can make the echo heights only approximately equal. Sophisticated modern machines have TCG control which can precisely compensate for the reduction of echo height with distance. See Fig. 6-C The time corrected gain is also called the time varied gain (TVG) or the swept gain. Other Controls Pulse repetition frequency (PRF) control : We saw earlier that the sequence of sending the pulse, receiving it, amplifying it and displaying it is repeated many number of times per second so that we get a constant trace on the CRT due to persistence of vision. It is the Clock circuit which controls the rate of repetition of this sequence. The number of times the pulses are emitted is called the pulse Repetition Frequency (PRF) or pulse repetition rate. Note : The pulse repetition frequency and the frequency of the ultrasonic waves are not to be confused with each other.

14

If PRF is increased the trace will become brighter, obviously. It must however be noted that certain combination of high PRF and high range setting can produce spurious indications called 'ghost echoes'. In some machines the PRF is not available as a control knob but is internally linked to the range selection knob. Normally a PRF of 100 to 1000 per second is employed. Monitor and gate controls : Some times one is interested only in certain section of the CRT picture. In such cases the section of the horizontal trace where we are interested can be selected or 'gated'. The gated section can be indicated either by a raised sweep line or a depressed sweep line or an intensified sweep line. Any echoes within the gate can be manually noted. Also echoes that exceed a predetermined height, called the 'threshold level', can be made to give out a visual and/or audible signal and also an electric output for recording on strip-chart recorders etc. Some equipment indicate this threshold level by a seperate trace which indicates the gate position, length and the threshold height. Generally the circuitry is so made that alarm indications are obtained when the echo exceed the threshold level. However, sometimes one would like to know when the indication falls below a threshold, as in the case of watching for loss of back-echo, and some equipment give this option also. Multiple gates are also possible. The gating system also helps to determine the region in which the time corrected gain (TCG) will be operable. The controls for adjusting the range to be gated and height to be monitored are called the monitor and gate controls. Marker Control The marker circuit gives small voltage pulses at a predetermined precise time intervals. The pulses are fed to the vertical deflected plates of the CRT and appear as periodic small pips on the CRT. The marker control indicates the time interval precisely and hence metal path can be read out easily from the CRT.

15

Chapter-6

EQUIPMENT CHARACTERISTICS
In the previous chapter we saw how an ultrasonic flaw detector functions and what controls are available to operate the UFD. In this chapter we shall explain the qualities or characteristics of a UFD which are of interest to us. The next chapter will tell us how to check for these characteristics. 1. Horizontal Linearity or Time-base Linearity of Sweep Linearity :

We have learnt that the range is set by setting just two echoes to their expected positions. This process assumes that the horizontal scale is linear i.e. equal distances represent equal thicknesses. For example, if the 5 th division on the horizontal scale represents 50 mm and 10 th division represents 100 mm, then the 2nd division should represent 20 mm and the 8th division should represent 80 mm and so on. If this does not happen, the horizontal scale is non-linear. The figure, shows a multiple echo pattern from a 10 mm thick material on a linear and non-linear scale. This non-linearity of time-base arises due to the non-linearity of the sweep voltage. Horizontal linearity is an equipment characteristic and is independent of the probe. Horizontal non-linearity cannot be rectified by the operator. However, a correction graph can be prepared showing scale divisions versus actual thickness. Horizontal linearity is important for measuring thickness or for accurate flaw location. 2. Amplifier Linearity or Vertical Linearity :

Vertical linearity or amplifier linearity refers to the height of echo 'A' being proportional to the signal height 'a' as provided by the probe, irrespective of the value 'A'. For example, assume that we have signals giving echoes of heights 40% and 80% of full scale height of CRT. Now, if the gain is reduced such that the first echo becomes 30%, the second echo should become 60% (i.e. maintain the same ratio of 2). If this does not happen, the amplifier is not linear. The equipment may show linearity upto certain height on the CRT. The vertical height of CRT where the linearity no longer exists is called vertical limit. In some equipment it may be observed that the echoes do not go beyond a particular limit even when gain is increased. This is called the saturation limit or saturation point. While using the UFD one should always operate below the vertical limit and saturation limit. Vertical linearity is also an equipment characteristic and is independent of the probe. Non-linearity cannot be corrected by the operator. This linearity is important when echo height ratios are to be recorded.

16

3.

Correctness of Amplifier Gain Control (dB Control) :

The gain control calibrated in dB which of course can be converted into linear ratios, if required. The calibration of the dB control should be accurate. For example, an increase of gain by 6dB should increase the echo height by a factor 2; and a decrease of gain by 20dB should decrease the echo height by a factor of 10. (See Table-1 on dB versus linear ratio). If this does not happen, the dB control calibration (i.e. markings) is not alright. The amplifier linearity along with dB control calibration accuracy permit us to reliably compare the heights of any two echoes. 4. Resolution (Spatial Resolution) :

Resolution is the ability to show distinct and well separated echoes from defects which are close to each other. Two cases arise : (1) defects which are separated by a small distance along the axis of the beam and (2) defects separated by a small distance perpendicular to the beam axis. The first case is called depth resolution (or axial resolution) and the second is called lateral resolution. Resolution is more dependent on probe than on the flaw detector. Lateral resolution improves with narrower beams. Depth resolution improves with narrower pulses. Narrow pulses are obtained by increasing frequency or increasing damping. We have already learnt about DEAD ZONE. It is obvious that the dead zone is nothing but NEAR SURFACE RESOLUTION (or the lack of it;). NEAR SURFACE RESOLUTION IS THE ability to resolve reflectors close to the entry surface from the transmitted pulse. Larger the dead zone, poorer the near surface resolution. High frequency and high damped probes improve near surface resolution as well. In addition, dual probes (which will be explained later) have better near surface resolution (or, in other words, less dead zone). The resolution explained earlier is also called FAR SURFACE RESOLUTION (as against near surface resolution). The instrument controls DAMPING and FILTERING can affect resolution. Higher damping and lesser filtering improve resolution. Lower pulse energy and lower amplification also help in bettering resolution. See Fig. 7

17

5.

Sensitivity :

Sensitivity is defined as the ability to detect small defects. Sensitivity is dependent on both the instrument and probe characteristics. Higher frequency lead to better sensitivity, in general. However, if the thickness range is high or the attenuation in the material is high, higher frequency may lead to reduced sensitivity. Damping decreases sensitivity. Higher pulse energy and higher amplification increase sensitivity. Use of suppression (reject) can prevent small echoes from being seen.

18

Chapter-7

CALIBRATION BLOCKS
For the successful accomplishment of the ultrasonic examination, it is necessary to have a good flaw detector and good probes. Most of these characteristics can be checked by using any arbitrary blocks of material. However, to ensure uniform practice in checking for these characteristics some standardised blocks, which are internationally recognised, are available. These are called calibration blocks. They can be used not only for checking the performance of the flaw detector but also of the probes. In this lecture we shall familiarise ourselves with the dimensions and features of some these calibration blocks. (I) THE I.I.W. CALIBRATION BLOCK (THE V-1 BLOCK)

This block originally designed through the efforts of the International Institute of Welding (IIW), is also popularly known as the V-1 block. See Fig. 8 We shall now briefly learn the uses of this block. This block can be used for checking the time-base linearity, the amplifier linearity, accuracy of dB calibration, depth resolution, near surface resolution (dead zone), for surface resolution, maximum penetration power, as well as some characteristics of angle probes such as beam exit point, beam angle etc. This block can also be used for setting the range (metal path) and sometimes can be used for setting up the sensitivity of the equipment for a testing assignment. a) Time-base linearity : Multiple echoes can be obtained from any of the thickness provided eg. 25 mm or 100 mm (position A or B in fig. shown) and time-base linearity checked as described in the previous lecture. Amplifier linearity : A multiple echo pattern is obtained from any convenient thickness and any two echoes are compared at different amplification level, again as described in the previous lecture. Near surface resolution (dead zone), (normal probes) : The probe is kept successively at position E, F, G and H and the nearest reflector, echo from which could be clearly resolved from the initial pulse is quoted as the dead zone. It should be noted that dead zone depends on probe-equipment combination and not on the equipment alone. Far surface Resolution (normal probes) : Depth resolution can be checked by keeping the probe at position C. If the three echoes from 85 mm, 91 mm and 100 mm are clearly separated on the CRT, the resolution is said to be good. This characteristic too depends on the probe-equipment combination.

b)

c)

d)

19

e)

Penetration power (normal probes) : The plastic insert in the V-1 block is made of a plastic material called perspex or lucite. This material highly absorbs ultrasonic energy generated by the given probe, all controls which affect the height of the echoes are kept such that echo heights will be highest possible. The number of back echoes obtained from the perspex insert, at this equipment settings, is a relative measure of the penetration power. This parameter also depends on the probeequipment combination. Range setting (normal probe) : To set the range to a given value in steel, multiple echoes can be obtained from a convenient thickness from the block, the thickness being selected such that more than one back echo appears on the CRT. Two of the echoes are to be adjusted to their calculated positions. The others will align themselves at their expected positions, provided that the time base linearity is good. Beam index (angle probe) : Beam index or probe index or the beam exit point is the point at which the beam leaves the probe in the case of angle beam probe. For checking the beam index, the probe is placed on the V-1 block to get the 100 mm radius echo. as shown in fig. The probe is moved forward and backward till the height of the echo from the radius is maximum. The beam exit point or the probe index corresponds to the centre of the centre of the 100 mm radius of V-1 block and it is to be marked on the probe. It is usual to measure and record the distance (in mm) of the probe index from the front edge of the probe.

f)

g)

h)

Beam angle (angle probe) : The beam angle is determined by using the angle graduations marked on the edges of the V-1 block. The probe is moved forward and backward till the reflection from the appropriate reflector is maximised. The probe angle is read off from the angle scale, corresponding to the beam index. Range setting (angle beam) : The angle probe is kept near the centre of the 100 mm radius quadrant of the IIW block. The first reflection obtained is from the 100 mm radius. The second reflection from the 25 mm radius after the reflection from the 100 mm radius does not occur at 100 + 25 = 125 mm, because the beam cannot be received at this wrong approach angle. The beam travels once again to the 100 mm radius and returns to the probe. The second echo therefore occurs at 225 mm. Subsequent echoes will also be at 125 mm intervals giving the sequence 100, 225, 350, 475 etc.

i)

The modified V-1 block with straight groove instead of the 25 mm radius groove is easier to calibrated with as it gives reflections at 100 mm, 200 mm, 300 mm etc. One must therefore note carefully which block one is using for range calibration for angle probing.

20

(II)

THE MINIATURE CALIBRATION BLOCK (V-2 BLOCK)

The miniature calibration block (popularly known as V-2 block) has the advantage of being small and handy. It has both some advantages and disadvantages over the V-1 block. See Fig. 9 Advantages : a) b) The block is small, light and handy so that it can be easily carried to inspection site. The block provides 12.5 mm thickness so that thickness less than 25 mm can be calibrated for with normal beam probes. It has quadrants of 25 mm and 50 mm radius so that ranges smaller than 100 mm can be calibrated for with angle beam probes.

c)

Disadvantages : a) b) c) There is no feature to check for the resolution of normal beam probes. There is no provision for checking the maximum penetration power. The block is thin and hence it is not reliable for checking the characteristics of large size angle beam probes. It is not suitable for calibration of large thickness ranges.

d)

It is to be noted that when the V-2 block is used for range calibration for angle beam, a series of echoes are obtained at 25 mm, 100 mm, 175 mm etc. If the beam is aimed at the 25 mm radius, and a series of echoes are obtained at 50 mm, 125 mm, 200 mm, etc. If the beam is aimed at 50 mm radius. (III) ASTM DISTANCE AMPLITUDE BLOCKS The ASTM distance amplitude blocks consist of a series of cylindrical blocks. These blocks help us to get the distance amplitude curve for a given flaw size. For this, blocks with same hole diameter (D) but with different metal paths (A) are selected and a distance amplitude graph constructed by recording the amplitudes of echoes from the holes at the same gain setting. See Fig. 10 (IV) ASTM AREA AMPLITUDE BLOCKS :

These blocks are identical to the distance amplitude blocks. For obtaining defect area viz. amplitude curves, blocks with same metal path (A) but having holes of different diameters (D) are chosen and the echo amplitudes are recorded at the same gain setting.

21

(V)

STEP WEDGES

The step wedges or stepped blocks provide a series of thicknesses which can be used for calibrations and checking of range and sweep linearity. Occasionally, the block is also used for checking resolution of the system by placing a normal beam probe at locations. See Fig. 11 (VI) HALF MOON BLOCKS

These blocks are simply semicylindrical blocks used for path calibration using angle beam probes. Obviously echoes will be obtained at distances of R, 3R, 5R etc. Refer Fig.12-A An improvement will be to include a step as shown in Fig. 12-B so that the block is useful for different metal paths. If many steps are incorporated, the block can be used a simple resolution check block for angle beam calibration. See Fig. 12-C (VII) OTHER CALIBRATION BLOCKS A variety of other calibration blocks are also specified and approved by different codes and standards. Each one has its specific advantage and limitations.

22

Chapter-8

THICKNESS MEASUREMENT
All along we have been implicitly talking about flaw detection. In addition to flaw detection, ultrasonic waves are also employed for thickness measurement, as well as for studying mechanical and metallurgical properties of material (e.g. strength, hardness, grainsize etc.). Thickness measurement is usually considered a part of ultrasonic testing, and hence we shall discuss it here. Two different techniques are made use of for thickness measurement by ultrasonic methods, viz. pulse-echo technique and resonance technique. Thickness measurement by pulse-echo methods : The ultrasonic flaw detector itself can be used for thickness measurement by pulseecho principle. Specially made thickness testers are also available for this purpose. (a) By using ultrasonic flaw detectors :

While discussing the principle of UT and the block diagram of ultrasonic flaw detectors, we learnt that the distance between the initial pulse and the reflected pulse, measured along the x-axis on the CRT, is proportional to the distance between the reflector and the entry surface. Thus, thickness of an object can be measured from the position of the back echo on the x-axis of the CRT. Since it is possible to read the position of echoes to an accuracy of 1% on the CRT, and horizontal linearity better than 1% is normally available, thickness measurement to an accuracy of 1% of the range is possible by using the ordinary flaw detectors. The accuracy can be further enhanced by two 'tricks' : (i) (ii) by observing a multiple echo (fig. shown) by using a lower range and bringing the back echo within the scale by zeroshifting. (This is called 'partial range setting'). Fig. shown is self-explanatory.

Of course, a combination of the two 'tricks' can be used to achieve the best results. It is possible to achieve an accuracy of 0.1 mm by employing these methods. However, there are two points to consider. Absorption and scatter may prevent getting clear multiple echoes; lower frequencies and lightly damped probes will help to overcome this problem. We need sharply rising echoes for accurate reading on the x-axis and this means we should use higher frequencies and highly damped probes. These two are conflicting demands and hence the most suitable probe should be selected for each application by trials.
23

Dead zone will prevent thickness testing of thin sections. The solutions will be high frequency probes; damped probes, observing multiple echoes etc. A special kind of probes called 'twin probes' or 'dual probes' were specially developed for testing thin sections and these probes can also be used for such applications. (b) By using thickness testers :

The use of the general purpose flaw detectors for thickness measurement can be cumbersome especially if a large number of readings are to be taken. Thickness testers specially designed for this purpose are available. These equipment have simpler electronics are cheaper, lighter and easier to calibrate and handle than the flaw detectors, Of course, their use is limited to thickness measurement. Two types of thickness testers are available. The first type make use of dual probes and they are useful for thicknesses from 2 mm onwards, and have normally an accuracy of 0.1 mm; these equipment are rugged and can be used on rough surfaces and absorbing material. The other type makes use of highly damped probes, Since energy output is small for highly damped probes, high amplification is used. These equipment can be used for thicknesses from 0.2 mm onwards, and have normally an accuracy of 0.01 mm. However, these equipment are delicate and it is difficult to use them on rough and highly curved surfaces or absorbing materials; for adapting these for such applications, the electronics of the equipment itself has to be modified. In the ultrasonic thickness gauges, the measurement of transit time is done by the 'integration method' or by the 'counting method'. In the integration method, the pulse applied to the probe initiates a current flow or voltage rise; when the reflected pulse reaches the probe, this current or voltage is stopped. The current or voltage reached during the transit time is proportional to the transit time and hence to the thickness and may be indicated either by analog meters (needle moving continuously over a scale) or by digital meters. In the counting method, the initial pulse starts a counter which counts the number of oscillations of an oscillator which is switched off by digital display. Many other methods of measuring transit time accurately are possible such as 'singaround method' etc. but these will not be discussed here. Thickness measurement by the resonance method : When 'continuous' ultrasonic waves are generated by the probe, the reflected waves will meet the incident waves and these two will interfere. When the thickness 'T' of the object happens to be an integral multiple of the half wave length of the ultrasonic waves, a phenomenon called resonance takes place. Standing waves are set up in the material under these conditions and higher current is drawn in the circuitry of the equipment. Thus by
24

varying the frequency of the ultrasonic waves and noting the frequency of the ultrasonic waves at which resonance takes place, thickness of the material can be calculated, As mentioned, the condition to be satisfied for resonance isT=n/2 = vn/2f (since, = v/f), where 'n' is an integer. When V is velocity of sound. Resonance will take place at a number of frequencies, corresponding to n=1, 2, etc. The frequency corresponding to n=1 is called the fundamental frequency of resonance and the other frequencies are called the harmonics. It is obvious from the above equations that the difference between any two consecutive harmonics is equal to the fundamental frequency. By continuously varying the frequency of the ultrasonic waves and noting the difference between two successive resonance frequencies, the thickness can be easily calculated. In equipment operating on resonance principle, the above procedure is carried out automatically and the thickness is simply read on a calibrated scale. Resonance technique for thickness measurement was evolved at a time when the pulse echo techniques could not give the required accuracy and could not be used for thin sections. However with the developments in electronics and probe manufacturing technology, thickness testers based on pulse echo technique have been able to match the performance of the resonance equipment. Thus resonance equipment have become obsolete and they no longer manufactured.

25

Chapter-9

GENERATION AND RECEPTION OF ULTRASOUND

Audible sound is produced in many ways such as by vibrating membranes or diaphragms (table, loud speakers), vibrating strings (violin, sitar) or by impact (tapping by finger on table or banging the door). High frequency sound (ultrasound) cannot be efficiently produced by these methods. A phenomenon called 'magnetostriction' can be used to produce ultrasonic vibrations but it is not useful in the megahertz range. For producing ultrasound in the megahertz range for UT, a phenomenon called PIEZOELECTRIC EFFECT is made use of. See Fig. 13 Only certain materials exhibit this characteristic of piezoelectric effect. When these piezo-electric materials are compressed, a voltage appears across their faces. If the compression is changed to dilation, the voltage still appears but the polarity changes. This is called PIEZO-ELECTRIC EFFECT. If alternating compressions and expansions (such as by a sound wave) are applied to the piezo-electric material, an alternating voltage of corresponding frequency is obtained. All piezo-electric materials also exhibit the reverse effect. If a voltage is applied to the material, it expands or contracts and if the polarity of voltage is reversed expansion becomes contraction and contraction becomes expansion. An alternating voltage will thus make the crystal vibrate with the frequency of the applied voltage. This is called REVERSE PIEZO-ELECTRIC EFFECT. Obviously REVERSE PIEZO-ELECTRIC EFFECT is made use of for generation of ultrasound and PIEZO-ELECTRIC EFFECT is made use of for receiving ultrasound. It is to be noted that the piezo-electric material is not made to vibrate at a desired frequency by applying an A.C. voltage of the same frequency and short duration is applied to the piezo-electric element in much the same way as table is tapped, or a sitar string is plucked or a tuning fork is struck. Such a pulse is known as a shock pulse or excitation pulse. The transducer element is activated by this pulse and it vibrates at its natural frequency. The natural frequency depends on the thickness of the piezo-electric element. In general, the frequency is given by f= V/2t where V is the velocity of sound in the transducer material and 't' is the thickness of the element. Thus, higher the frequency, thinner is the piezo-electric element. It should be kept in mind that this frequency may get altered by the manner is which the element is incorporated in the search unit as well as by using the Damping control of the flaw detector. Quartz was one of the earliest piezo-electric materials used. Quartz crystals have a hexagonal shape and have many axis of symmetry, the crystal contracts and expands on
26

applications of an a.c. voltage as shown in figure above. Thus longitudinal waves are generated and such a slice of quartz crystal. If, on the other hand a slice is cut perpendicular to the y-axis the crystal vibrates in shear mode (figure shown) and such a slice is called ycut crystal. For various reasons, quartz is no longer popular as a generator/receiver for ultrasound in UT. The materials most commonly used are not single crystals but produced by powder metallurgical methods. However the piezoelectric element is still often called the 'crystal', because of the fact that quartz 'crystals' were used in earlier days. For every piezo-electric material, there is a temperature above which the material loses its piezo-electric property. This temperature is called the Curie Temperature. Most of the modern transducers are polycrystalline in nature and the damage due to heating cannot be reversed by cooling. All piezoelectric materials are not similar in their properties. Some are good transmitters, some are good receivers, some have good physical/chemical stability etc. A comparative table is given below. It must however be noted that this table is to be taken only as indicative. The characteristics of a piezoelectric element can greatly vary depending upon the method of manufacture and they can get further modified by the way they are incorporated in the search unit or probe.

Piezoelectric Material:
Material as transmitter @ VP P G VG as receiver @ F VG P P as transceiver @ P G F G Resolution @ G G F P Stability Curie Temp. 5750C 750C 1200C 3600C

Quartz LiSO4 BatiO3 PZT

VG VP G G

@ : These characteristics can be considerably modified by probe construction. VG -Very good G - Good F- Fair P - Poor VP - Very Poor

27

Chapter-10

NORMAL AND ANGLE BEAM PROBES

The probe is the most critical part of the ultrasonic test system. Its abilities and limitations define all aspects of UT, from instrument design to test specifications. The main types probes are 1. NORMAL BEAM PROBE or STRAIGHT BEAM PROBE : The beam is parallel to the normal to the surface on which it is incident. (In other words the beam is perpendicular to the in surface on which it is incident ). Refer Fig. 14 2. ANGLE BEAM PROBE The beam is at some angle, say 45 degrees, to the normal to the surface. See Fig. 15 3. DUAL PROBE A normal probe but with seperate crystals for transmitting and receiving; main uses of this type of probes is detection of flaws close to the surface and thickness testing of thin sections. 4. IMMERSION PROBE The probe dips into water in which the job immersed. There is no probe-to-job contact and the water acts as the couplant. Curved lenses can be fitted to the probe to get the beam focussed. 5. SPECIAL TYPE OF PROBES Mosaic, array or matrix probes; wheel probes for plate testing etc. Construction of normal probe : The figure 14 & 15 show the cross-section of a typical normal probe. The piezoelectric element ( or the crystal) is coated on both sides with thin conducting layers. The contact surface conducting layer is internally connected to the probe case; this in turn gets 'earthed' or 'grounded' through the cable and the instrument. The metallic conducting layer on the other side or top side, is connected to the central contact of the probe connector (either directly or through an electronic network). The voltage pulse is applied from the instrument through the cable to this point. The crystal and its metallization is rather delicate and has to be protected against damage and wear when used on rough workspaces. There are several protection methods.
28

1.

Hard Wearface : Probes of this type are called hardfaced probes. A thin layer of very hard material such as ruby or alumina is bonded to the metallized crystal (at the manufacturing stage). Such probes give a long life on smooth surfaces; these probes can yet stand quite some amount of abrasion from moderately rough surfaces. The wear face may be somewhat brittle. Wear Membrane : Wear membrane probe are designed to be used with a thin replaceable membrane which is stretched over the probe contact surface, with some oil in between. Such probes should never be used without the membrane except for an occasional direct light contact and even in such case it should not be slid over the surface; otherwise, the metallizing or even the crystal will be permanently damaged. Wear membrane protection is more suited for low frequency probes and improves contact on curved and rough surfaces. The membrane, made of a special grade of plastic, can be replaced in the field quite easily. Wear Caps and Delay blocks : Occasionally a rather thick (2 to 3 mm approx) plastic wear cap is used. This would adversely affect near surface resolution but can be used on very rough surfaces. A very long wear cap would actually be called a 'delay block' and would give quite good-near-surface resolution (i.e. less dead zone), but test range in say, steel would be limited to a maximum of approximately twice the plastic wear cap length.

2.

3.

Also note that some old quartz probes were used without any protective layer, quartz being quite hard and wear resistant. The ground electrode for such probes is the test piece itself, contact being maintained by a spring loaded ring around the probe front. As ultrasonic probes have to be damped to achieve better resolution, a damping body is bonded to the back face of the crystal. This damping body, called 'backing member', should ideally absorb the vibrations of the back surface of the piezoelectric element. Also the energy received by the damping body should be totally absorbed, as otherwise they will be reflected back to the crystal. To meet these critical requirements, the damping body is usually made of plastic or rubber containing powdered heavy metals. Finally the probe must be reasonably water-tight to avoid damage from the various liquids used as couplants. The case is usually made of stainless steel or aluminium to avoid corrosion from couplant as well as for mechanical strength. Probes are usually specified by size (diameter or sides of rectangle or square) and frequency, e.g. 2MHz, 24 mm dia probe. Note that the diameter refers to crystal diameter and not case diameter; the case obviously will be larger. These two parameters are invariably written on modern probes or are at least available from the probe specification sheets. A few manufactures also give some arbitrary coding detailing the amount of damping. This coding is not standardized between various manufacturers.
29

Construction of Angle Beam Probes : Angle beam probes are used extensively in UT, probably in about 50% of all UT work, considering the amount of welds being tested by ultrasonics. An angle beam probe essentially a normal beam probe mounted on the plastic wedge. This causes the beam to be incident on the job at an angle. The beam travels in to the job at an angle larger than the incident angle. The cause of this phenomenon (called refraction) and the fact that this beam within the workpiece is of a different character (shear waves). In one common type of a probe, the crystal is directly bonded to the wedge on the one side and to the backing member on the other. The plastic of which the wedge is made is usually perspex (also called plexiglas or lucite) and it is a clear transparent material. Not all of the beam enters the workpiece. Quite a good amount of energy is reflected obliquely inside the plastic wedge. This energy, if it reaches the crystal again, will produce disturbing echoes on the CRT. So, the entire thicker end of the wedge is bonded to a material which absorbs sound wall. The bonding surface is frequently serrated for better absorption of the sound and vulcanised rubber is a common material used for this purpose. The entire assembly of crystal with damping body, wedge with the sound absorber and connector is enclosed in a steel case. Also available are 'removable wedge type' angle beam probes. These permit crystals of different frequencies to be used inter-changeably with wedges of different angles. Removable wedge type is convenient because it is comparatively cheap to make wedges for a new or non-standard angle or to match curved surfaces of the workpiece. The crystal assembly is clamped to the wedge by screws. Crystals meant to be used with removable wedges should not be used as normal beam probes since they are not wear resistant and may perform ultrasonically well only on plastics. For the latter reason it is not advisable to use a normal beam probe on a wedge to convert it to an angle beam probe. Note that the angle beam probe sends shear waves into the workpiece which travel with a velocity of 3280 m/sec. in steel. For angle beam probes, change in test material not only means a change in velocity but also a change in the (refracted) beam angle. In addition to the size and frequency, angle beam probes are to be identified by an angle also. The angle marked on the probe is not the incident angle in the wedge but the refracted angle in steel. If the material is different from the steel the angle also will change and this can be calculated. Standard angles available are 30, 35, 45, 60, 70 and 80 degrees.

30

Due to the contact surface being of softer plastic, angle beam probes get more easily worn out. A uniform wear has no noticeable effect on probe performance; only the point at which the ultrasonic beam leaves the wedge (the 'beam exit point' or 'beam index' or 'probe index') will slightly change. But, non-uniform wear will change the wedge angle and hence beam angle will change in addition to the beam index. This has to be checked periodically. Testing on curved surfaces may make the wedge also curved and it may not make good contact on flat surfaces. In such case the wedge should be ground flat. For reasons mentioned above it is necessary to check the beam index and the beam angle of a probe periodically. This has been discussed in an earlier lecture. Skip and path distances : When testing plates, or welds in plates, the angle beam get reflected between either surfaces of the plate and thus shows a zig-zag path. In the above figure, the distance AC is called the (full) 'skip distance' and the distance ABC is called (full) path distance. The skip distance and path distance are determined by the thickness of the plate (t) and the beam angle () Full skip distance = 2 x t x tan Full path distance = (2 x t) /cos Flaw location with angle beam probe : If a flaw is encountered by the beam before the half skip as shown in figure 16, the projected distance (X) of the flaw from the beam index and the depth (Y) of the flaw from the surface are given by : X = b sin Y = p cos , where 'p' is the beam path as seen on the CRT. If however, the defect is beyond the half skip, the formulae get a bit complicated. It is easier to draw to scale a beaming diagram and mark the beam path on it and thus read the defect position from it. See Fig. 16 Skip and path distances in pipes : The axial (or longitudinal) scan on pipes is identical to the case of plates discussed earlier. However in the circumferential scanning, the skip and path distances are not given by simple formulae and they found out by drawing to scale a beaming diagram. See Fig.17 For circumferential scan, there is an angle at which the beam touches the ID tangentially. This angle is called the limiting angle (L). If the angle used is more than this
31

limiting angle, the beam will miss a region near ID. Therefore angles less than limiting angle should be used during circumferential scan. Limiting angle is given by the formula Sin L = ID/OD = Ri/R0 Connectors : We have seen that the probe cable is attached to the probe and the instrument by connectors. There are many types of connectors. Some common types are : (i) Lemo 0 (ii) Lemo 00 (iii) BNC (big and small) (iv) Microdot (big & small) (v) tuchel (vi) UHF and (vii) subvis. If for some probe or instrument connector, a suitable cable is not available, we should obtain suitable adaptors e.g. Lemo-to-BNC. Cables : Ordinary electrical wire would not work for UT probe cables because the ordinary wire will pick up electrical interference due to high frequency machines (such as welding, drilling etc.) working nearby. The special cable used as probe cable is called 'co-axial cable' in which the electrical voltage is applied to a central conductor which is surrounded by an insulation is surrounded by a conducting wire mesh which is earthed; this earthing wiremesh is further covered by an insulation. Care of probes and cables : Plug in connectors without excessive force or twist; each connector has its own way of connecting. Cables should not be bent too much and should be cleaned to remove couplant after use. Avoid damage to cables from sharp edges of the material being tested. Do not pull the cable off from the equipment; instead, unconnect the cable by holding at the connector. Replace the cables when they fail. Probes are delicate; dropping a probe may be the end of it, especially if it is a high frequency probe. Avoid excess pressure or rubbing on the probe. Use the probe as it is intended to be; probes with protective membranes should not be used without the membrane; angle probes to be used with replaceable wedges should not be used without the wedges (as a normal beam probes). Immersion probes should not be used as contact probes. Probes should not be used on surfaces hotter than the temperature for which they are intended. Probes with a length of cable permanently attached should be handled with care since the cable cannot be easily replaced if damaged. L=Sm-1 Ri/R0

32

Chapter-11

DUAL PROBES
We have seen that the transmitted pulse (or the initial pulse or the main bang) occupies some distance on the timebase and therefore any echo within this region cannot be clearly seen. This distance (in mm of material), within which defects cannot be detected is called the dead zone. Dead zone can be eliminated if the transmitted pulse can be prevented from entering the amplifier. This is possible if a seperate receiver is employed, as shown in figure 18. Another way of explaining the dead zone is that the crystal vibrates for some definite time and during that time it cannot act as a receiver. If we can use a seperate receiver then, echoes can be received even the transmitter is active. Handling of a transmitter-receiver combination will be convenient if the seperate transmitter and receiver are incorporated in a single housing. Such probes are available and they are called DUAL PROBES, TWIN PROBES, DOUBLE CRYSTAL PROBES, T-R PROBES etc. The transmitter and receiver should be obviously of the same frequency. They must be slightly inclined towards each other (see figure shown) if the reflected beam is to reach the receiver. This angle which the crystals make with the horizontal is called the 'roof angle'. Delay blocks usually of perspex are added to both the crystals. If the reflection from the probe front surface reaches the receiver, it will give rise to spurious signals. This is called cross-talk. To prevent cross-talk an acoustic barrier, through which acoustic waves cannot pass, is kept between the two. In spite of the acoustic barrier, a small-cross talk echo occurs via. the front surface of the probe. This, being small, is seen on CRT only at very high gains. It is sometimes useful for identifying the 'zero' for range calibration. Since the probe has definite physical dimensions, it still cannot receive signals from defects very close to the surface (figure above). Hence even dual probes have dead zone but the dead zone is much less than in the case of single element transducers. Dead zones as small as 2 mm are achieved in the case of dual probes. It must be noted that due to the roof angle the beam enters the material at a small angle. Due to this and due to the finite dimensions of the crystals, echoes outside certain range of depth are not received well by the probe. This is called the 'effective range' of the dual probe and this effective range depends upon the roof angle and the size of the crystals. The effective range, which is normally a disadvantage, sometimes an advantage. If the material is highly scattering, the scatter from the entire thickness will appear on the CRT in the case of a single element probe, whereas in the case of the dual probe noise from the effective range alone will enter the probe. Therefore, noise is reduced on the CRT
33

and thus the signal-to-noise-ratio (S/N ratio) is increased. Certain precautions are required when using the dual probe. The delay blocks should not be unevenly worn-out. Large wear-out of the delay blocks, will alter the effective range. While using the dual probe on curved surfaces, the dividing line between the two sections of the probe should be tangential to the curved surface. Another precaution is regarding the range calibration. In ordinary probes multiple echoes are used as reference points for range calibration. For dual probes multiple echoes should not be used as the distance between zero and first echo is not equal to the distance between first and second back echo. Hence, two seperate thicknesses are to be used for calibration. Dual probes were developed to overcome the problem of dead zone. Of late, highly damped probes with narrow pulse width have seen developed, and these are slowly replacing dual probes. Angle beam dual probes : Dead zone is not a great problem in angle beam testing. This is because large part of the dead zone goes in the perspex wedge. Also the angle probe can be moved back to see the same defect at a greater distance. Thus angle beam dual probe was never thought of till recently. However, the improved S/N ratio of the dual probe is useful in the case of highly scattering material like austenitic stainless steel welds and angle beam dual probes are coming into use for this purpose.

34

Chapter-12

PULSER AND RECEIVER

Pulser : The electric pulses emitted by the pulser excite the piezo-electric crystal. Higher the voltage of these pulses, large will be the amplitude of crystal vibrations and thus more energy will be the transmitted in the form of ultrasonic waves. However, this will make the crystal vibrate for longer time and consequently the ultrasonic pulses will be wider. This will mean poorer resolution (both near-surface and far-surface resolution). Therefore it is advisable to operate at lower pulse energies and use higher amplifications to achieve the required sensitivity. However, it is advantageous if the equipment gives the option of high pulse energy operation because this will be useful for large thickness ranges and/or highly absorbing material. Receiver (Amplifier) : (a) (b) Linearity : We have seen already the importance of linearity of the amplifier. Total gain : The higher the gain of the amplifier, the better the sensitivity as smaller echoes can be amplified to give observable echo amplitudes on the CRT. Frequency response : The amplifier can have broad-band or narrow band response.

(c)

Broad-band amplifier has the advantage that the broad-band pulse emitted by the highly damped probes will be fully amplified. Narrow band amplifier has better sensitivity within its band and also reduced noise since frequency components outside the narrow band will be automatically filtered. It is preferable to have both broad band and narrow-band options.

35

Chapter-13

PROFILE OF ULTRASONIC BEAM

A circular piezo-electric transducer does not give rise to a cylindrical beam of ultrasonic energy. The beam actually converges upto a certain distance (N) and then diverges. See Fig. 19 The region from the probe upto the 'neck' of the beam where it starts diverging is called the NEAR ZONE or FRESNEL ZONE. The distance 'N' is given by the formula N=D2/ 4 he region beyond the near zone is called the far zone or Fraunhofer zone. In the far zone the beam spreads out or diverges and it appears as if the beam comes from a point source at the centre of the crystal with a semi-cone angle given by Sin =K / D, where K is a constant equal to 0.51 or 0.87 depending upon whether beam boundary is defined as the region where the intensity is half (i.e. minus 6dB) of the intensity at the axis. If one studies the intensity variation along the axis, it is seen that beam varies irregularly upto N and then decreases regularly. Thus, defects of same size will give different echo amplitudes within 'N' and the echo will decrease in a regular fashion in the far field. Thus flaw evaluation in the near zone can load to errors. In the far field, the intensity varies inversely as the square of the distance (inverse square law). Similarly if the intensity across the beam (i.e. perpendicular to beam axis) is plotted, one observes irregular variations in the near zone but a smooth pattern in the far. Since the intensity across the beam is not uniform, all the points within the beam are not scanned with same sensitivity. Therefore, it is customary to overlap scan areas between successive scans. It must be noted that damped probes do not operate in a single frequency or a narrow frequency range but they emit a wide range of frequencies. Therefore the above discussions will not hold good for such probes. The beam shape, if needed, should be determined experimentally for such probes. Note 1 : It is also customary to simplify the profile of the beam as shown. Note 2 : Since the beam shapes discussed above are based on different assumptions and approximations by different authors the value of 'K' in the beam spread formula is also quoted differently. The numbers given are to be taken only as nominal.
36

Chapter-14

WAVE PROPAGATION IN MATERIAL

Passage through matter : Absorption, scatter and attenuation : Beam divergence is not the only reason for the reduction in intensity of the beam as the distance increases. In a real material, a sound wave also continuously loses a part of its energy through conversion into heat and this called (pure) ABSORPTION. In addition a part of the sound wave is scattered from microscopic interfaces in the material and this is called SCATTER. The combined loss of energy due to absorption and scatter is known as ATTENUATION. In UT, it is common to use both the terms absorption and attenuation to refer to the combined effect. The sound attenuation increases with an increase in frequency. Attenuation also depends upon the material (absorption is different for different materials) and its metallurgical structure (this affects scattering). Intensity reduces exponentially with distance due to attenuation. The net decrease in intensity with distance is a combination of decrease due to beam divergence (which follows inverse square law) and that due to attenuation (which follows an exponential law). Normal incidence at an interface : reflection and transmission When sound is incident normal to (i.e. perpendicular to) an interface between two media, it is partly REFLECTED and partly TRANSMITTED. The ratio of sound energy reflected (Er) to the sound energy incident (Ei) is called the reflection factor R. The ratio of the transmitted sound energy (Et) to the energy incident (Ei) is called the transmission factor T. This factors depend upon a property of the material called 'ACOUSTIC IMPEDANCE' (Z). Acoustic impedance of a material is the product of its physical density () and the velocity (V) of sound in the medium. The z= x V. The factors are shown. See Fig. 20 Oblique incidence : reflection and transmission and mode conversion : When the ultrasonic beam is incident obliquely (i.e. at an angle) to the interface, the reflected wave is also at an angle to the surface. The angle of incidence (A) is equal to the angle of reflection (B). However, the transmitted wave undergoes an abrupt change in direction and this phenomenon is known as REFRACTION. See Fig. 21 The angle of refraction (C) is given by the formula (known as snell's law): Sin A/ Sin C = V1/V2 where V1 and V2 are the velocities of the sound waves in the first medium and the second medium respectively.

37

We are familiar with such phenomena with respect to light. However the reflection and refraction of sound waves are not as sample as mentioned above. It is observed that the reflected wave consists of two components, one of longitudinal mode and the other of shear mode. The transmitted (or refracted) wave also similarly consists of two components. The angles of reflection and refraction are given by the same Snell's law provided the appropriate pair of velocities are substituted. This phenomenon of conversion of one mode of waves into another due to reflection or refraction is called MODE CONVERSION. We have seen that for angle beam inspection the longitudinal sound beam from a crystal is incident through a plastic (perspex) wedge at an angle to the material surface. Since mode conversion will introduce two waves of different velocities and angles into the test material, the results will be confusing. Therefore we would like to eliminate one of them. Since the velocity of longitudinal waves is greater than the velocity of transverse waves, the refracted angle of longitudinal waves is always greater than that of transverse waves. Now if the incident angle is increased, the refracted angles (of both refracted waves) will also increase and at a particular incident angle the refracted angle of longitudinal wave will be 90 degrees and thus only shear waves (transverse waves) will be present in the second (refracting) medium. The corresponding incident angle is called the FIRST CRITICAL ANGLE. On further increase of the incident angle, refracted angle of shear wave will continue to increase and will reach 90 degrees. The corresponding incident angle is called the SECOND CRITICAL ANGLE. The shear waves travelling at right angles to the normal (i.e. travelling along the surface) are called the surface waves or Rayleigh waves. For angle beam scanning the incident angles should be between the first and second critical angles. The corresponding refracted shear wave angles in steel are 32 degrees (approx.) and 90 degrees. Due to beam spread surface waves may be generated even when the refracted angle for shear waves is about 80 degrees. Therefore beam angles of 35 degrees to 70 degrees are recommended. A 45 degree beam is the ideal one. It should also be noted that due to the phenomena discussed above, shear waves are (almost) exclusively used in angle beam testing and many times angle beam testing and many times angle beam testing and shear wave testing are used as equivalent words. The partition of energy between the four beams ( two reflected and two refracted) can not be expressed by simple formulae. Mode conversion can give raise to unexpected results. A shear beam of 30 degree (approx) gets strongly mode converted into longitudinal waves. Mode conversion can also occur because of the beam angle changing due to geometry.

38

Reflection by 'large' and 'small' reflectors : The interface phenomena of reflection etc. discussed above are true for interfaces much larger than the wave length. When the reflector, such as a defect, which is of the order of the wave length is encountered, the sound beam does not get reflected geometrically. Rather the sound bead is reflected as a wide divergent beam. As the defect size decreases, the divergence of the reflected beam increases. Therefore the amount of energy returning in the direction of the incident beam (or in the expected direction in the case of angular incidence) is small. Thus it becomes difficult to detect defects smaller in size than the wavelength. To detect a defect few mms in size the wave length of the ultrasonic wave should also be of the same order. This is why one has to use sound waves whose frequency is in the range of MHz, and not lower frequencies. Thus we need 'ULTRA' sonic waves and not audible frequencies for testing.

39

Chapter-15

INTERACTION OF ULTRASONIC WAVES WITH MATERIAL

Beam shape : =V /f. Therefore, for a given frequency of ultrasonic vibration, wavelength is different in different materials. We already know that wavelength affects beam shape. (Since nearfield N=D2 / 4 and divergence angle =Sin-1 (k / D) Thus beam shape is a function of the material for a given probe (i.e. for a given diameter and frequency of transducer). Attenuation : We already know that attenuation (both absorption and scatter) is a material characteristic. Any inhomogeneity in a material increases scatter. Scatter may arise due to large grain boundaries, grain boundaries precipitations, segregations, distributed porosity and inclusions, coring of the grains etc. Scatter becomes significant when grain size /10. (This is only a thumb-rule). Effect of anisotropy : Anisotropy is a property due to which the velocity is different in different crystallographic directions. In a randomly oriented grain structure, when sound enters from one grain to another, there is a change in velocity and a change in acoustic impedance. This means more scatter due to reflection and refraction in grain boundaries. More the anisotropy more is the scatter. Metallurgical history of material and attenuation : Castings : Usually have large grain size, porosity, segregations, cored grains, grain boundary precipitations, inclusions, etc. Hence castings generally scatter more and hence one has to use lower frequencies. High scatter and use of lower frequencies reduce achievable sensitivities. Wrought Products : (like rolled/forged/drawn/extruded products) have lower grain size; casting defects get reduced or eliminated in wrought products; inhomogeneities in composition also get reduced. Consequently, the scatter is less and wrought products can be tested to high sensitivities. The working process may elongate the grains in one direction. Thus scattering will be more when ultrasonic beam is perpendicular to these elongated grains, and scatter will be less when the beam is parallel to the elongated axis of the grains. Welds : Typically should show cast structure. However, due to the small volume of molten metal and due to fast cooling, the grains in weld are sufficiently small. Hence scatter in a weld (such as carbon steel welds) is not significant. However, there are certain materials like austenitic stainless steel and Ni-based alloys whose welds show large grain size. Added to this they have high anisotropy. Hence welds of these material (unless section thickness is very low, a few mm) scatter heavily and UT may sometimes becomes impossible.
40

15.1

FLAW EVALUATION

When a flaw is detected, one would lime to know the characteristics of the flaw so that decision can be taken regarding its acceptability. These characteristics are : (i) location (ii) size (iii) shape (iv) type and (v) orientation. I. LOCATION

When a flaw is detected, its location is immediately known from the beam path. Accuracy of location within 1% of scale range is possible. Inaccuracy can creep in due to beam spread. For example, during angle beam inspection of thin sections, it may not be possible to distinguish whether the defect on the top surface, bottom surface or mid-section. II. SIZE

The size of the flaw is easily determined if the defect is much larger than the beam, by moving the pipe and marking the boundary where defect echo disappears (or reduces to a fraction of maximum height). However, it is difficult to size defects smaller than the beam by the above method. The amplitude of echoes is the only parameter available for this purpose. For a given probe, the echo from a given size defect will vary with distance. This variation can be studied with artificial defects (e.g. flat bottom holes) drilled to different distances. The graph of echoamplitude Vs distance for a given defect size is called the DAC (or distance amplitude correction) curve. Similar curves will be required for other sizes of defects. Thus we will have a set of DAC curves. These set of curves will be different for different probes. The curves will also vary for different material. Manufacturer's of certain equipment give what are called DGS (Distance-Gain-Size) scales which are nothing but DAC curves for specific probes. These curves are very handy. However the following should be kept in mind : 1. These curves assume perfect reflection from defects (such as from a flat bottom hole). They assume a fixed amount of low attenuation (corresponding to wrought carbon steel). If attenuation is different in the workpiece corrections are to be applied. They assume a very highly damped probe. They are graduated for carbon steel. Vertical and horizontal linearities should be good.

2.

3. 4. 5.

41

It should also be borne in mind that what we get from DAC curves are sizes of flat bottom holes (FBM) whose echoes are equal to the echo from the flaw considered. Reflectivity of a defect opening to the surface (cracks, seams, laps etc) is more than that of an embedded crack. Hence flat bottom holes (FBH) does not represent such defects. Also elongated defects and circular defects (FBH) of equal area may not have equal reflectivity. Hence it is customary to make reference blocks containing artificial defects of required type. Reference blocks, eliminate the need for corrections due to material variations. Notches, side drilled holes (SDH) and flat bottom holes are used as artificial defects. Notches represent surface opening defects like cracks, laps etc. Flat bottom holes represent embedded defects like inclusions, laminations etc. side drilled holes represent linear defects inside the material. Of these SDH are easier to fabricate and the same defect can be used as reference standard for both normal beam and angle beam. Since there is a variety of artificial defects and also their location is variable, one should use the reference standard specified in applicable codes or procedures. When artificial reference standards are used, the defect size is not calculated but is mentioned as a relative quantity; e.g. as bigger or smaller than the reference defect. III. SHAPE

Again, if the defect is very large compared to the beam one can know to shape. However this is true only for planar defects. For three-dimensional defects and detects smaller than the beam, information on shape is a question of judgement. The following guidelines will be useful. An irregular shaped detect gives an irregular shaped echo. (Remember to put off suppression and filtering if echo shape is to be studied). Echo from a planar defect will vanish if the beam is tilted. Echo from volume defect will not appreciably change. IV. TYPE

The type of defect is judged by the size, shape, orientation and location as also by the shape and behaviour of the echoes. Planar defects give sharp echoes and inclusions; bunch of pores etc. give irregular shaped echoes. V. ORIENTATION

Orientation is determined from the beam orientation at which maximum echo is obtained.

42

Chapter-16

SELECTION OF TEST PARAMETERS

So far, we learnt about -(i) (ii) (iii) (iv) (v) equipment characteristics probe characteristics physics of wave propagation in a material interaction of material characteristics with ultrasonic waves, and interaction of defects with ultrasonic waves.

We shall now consolidate our knowledge by discussing how to choose test parameters for a given testing assignment. When a testing assignment is given to us, we should ask ourselves the following questions : (A) What is the chemical/metallurgical composition of the material ? Normally for iron based alloys, no specific problems are faced with respect to composition. Austenitic stainless steel is an exception. This is because austenitic stainless steel castings and welds have very large grain size (sometimes as big as a cm). Wrought austenitic stainless steel also can have large grain size. Large grain size means high scatter. Copper, lead, copper alloys (e.g. brass) and nickel alloys (e.g. monel) also give problems due to absorption/scatter. Plastics absorb ultrasonic energy heavily; composites may show both absorption and scatter. What is the fabrication history of the material ? Castings are generally inhomogeneous, and have large grain size. Castings which have undergone homogenising heat treatment will be more homogeneous and therefore less scattering. Grey cast iron in which graphite is in the from of flakes gives more scatter than spheroidized cast iron where graphite is in spherical form. Wrought products (rolled, forged, drawn, extruded) have finer grains and are more homogeneous. The grains may get elongated in one direction and hence scatter may be more in one direction than in the perpendicular direction. Casting defects get eliminated or get elongated. The working itself can cause new defects.
43

(B) -

Welds in principle are castings but due to thin sections and faster cooling grains are usually sufficiently small. (Exceptions are austenitic stainless steel Ni alloys etc). Welds have their own defects. Powder metallurgical products, if well made, can be as good as wrought products. If not made well, can be as scattering as castings. Machining, grinding etc. also can introduce surface defects, especially in high strength material. What is the size and shape of the material ? In very thin sections, it becomes difficult to resolve echoes which will be closely spaced. Very high thickness may absorb heavily. In general, proper parameter selections can be made to test sections as thin as 0.1 mm to as thick as 1000 mm. Shape can highly restrict UT. Very high curvatures (e.g. 10 mm OD) may make conventional manual scan ineffective complex shapes such as gears, crankshafts etc. may make UT ineffective or laborious. Sometimes only a part of the component may be amenable for UT. Surface roughness also affects UT. What is the type and location of defects ? Planar defects are very easily detected compared to three dimensional defects. Defects breaking open on surfaces are also more easily detected than defects embedded inside material. Defects close to surfaces may need different selection of parameters than defects inside material. What is the sensitivity required ? Obviously gross defects are easily seen compare to small defects. Adverse conditions of composition, fabrication history, shape and size of material etc. may limit the sensitivity achievable.

(C) -

(D) -

(E) -

Now let us see what are the parameters that a UT inspector can manipulate and how he can manipulate them to get best results for different combinations of material characteristics discussed above. I. 1. EQUIPMENT PARAMETERS Good horizontal linearity is important if flaw location should be accurate or accurate thickness testing is needed. Vertical linearity is a must if echo heights are to be determined as a percentage of
44

2.

some reference height. If however recording conditions mentions "more than or equal to the reference echo height", then vertical linearity is not very important. However, one should always operate below saturation limit. 3. Thin sections need good resolution. However resolution depends on equipmentprobe combination. Equipment having pulse energy and/or damping controls permit manipulation of resolution. Low pulse energy and high damping improve resolution. The above (para 3) holds good when high penetration or application to highly absorbing material is concerned. High pulse energy and low damping improve penetration and sensitivity. If high damping through the equipment is to be applied or if highly damped probes are to be used, broad-band amplifier will be advantageous. PROBE PARAMETERS Crystal : Comparison of crystal characteristics was done during an earlier lecture. Since these characteristics will depend on method of manufacture and can be altered by probe construction, the operator cannot have strong criterion for choosing one crystal from another. Many times the manufacturer does not give information about the crystal material. Size : Generally selection of probe size is a question of convenience. Large probes speed up scanning. Small probes are useful on curved surfaces or restricted spaces. It should be kept in mind that smaller probes give more divergent beam and larger probes have longer near fields. Frequency : Higher frequencies can detect smaller defects since wavelength is small. However since absorption and scatter increase with frequency a compromise must be made. Also higher frequencies give less divergent beam and also better resolution. Broad-band (damped) probes improve resolution but sensitivity may become low, due to low energy output. 4. Single crystal or dual crystal probes : Dual crystal probes have low dead zone. However each dual crystal probe has detect flaws. Highly damped probes are slowly replacing dual probes. Dual crystal angle probes are sometimes used for their improved signal-to-noise ration. Beam angle : Ultrasonic beam must be incident perpendicular to the defects. Defects parallel to scanning surface are therefore detected by normal beam. Due to beam spread and other reasons, defects oriented as much as 100 to the beam will still be
45

4.

5.

II. 1.

2.

3.

5.

detected. If geometry does not demand other angles, 450 is most convenient. Three dimensional defects (e.g. inclusions, porosity etc.) are detected by both normal beam and angle beam. III. SCANNING DIRECTION

Scanning should be done such that the beam hits defects perpendicularly. For example, circumferential defects in pipes will be detected by axial scanning and axial defects by circumferential scanning. IV. TIME OF INSPECTION

For a given product, UT can be done at different stages of manufacture. It may be advantageous to inspect the material at one stage than the other. For example, casting are better tested after homogenising heat-treatment. If grooves, steps, teeth etc. are to be made, it is better to carry out UT before such operations to avoid problems of geometry. Since improper heat treatment can cause cracks in certain cases, UT should be done after heat treatment. Bars are better tested before boring, since dead zone can be overcome by scanning from opposite direction (1800 apart) in solid bars.

46

Chapter-17

ACCESSORIES
So far we learnt about the basic features of UFD and probes. For easier manual scanning, for better data collection or to automate/mechanize the entire testing, certain accessories are used. In this lecture we shall learn something about the common accessories used in UT. Accessories and modifications to the probe : To facilitate manual scan, spring loaded probe holders (to ensure uniform contact), or curved shoes (to fit the contour of the job) may be used. To prevent wear and tear to the probe, to improve near surface resolution or to enable semi-automation, sometimes a column of water is used as couplant. To maintain the water column as assembly is called a 'bubbler' or 'squirter'. To make plate testing easy, the probe can be attached to an assembly which looks like a lawn mover so that the UT operator can walk on the plate with this 'plate tester' to test plates. A series of probes can be fixed in a holder and they can be excited simultaneously or in sequence. This can speed up testing and/or minimise the need for probe motion. Such probe systems are called probe arrays, or probe mosaics. Operating the probes sequentially is called 'multiplexing'. Strip chart (i.e. paper) RECORDERS can be connected to the UFD so that defect indications are recorded. For this the equipment should have a monitor or gate facility (about which we have learnt earlier). The equipment gives voltage output proportional to height of echoes within the preselected gate area. This is useful only with automated or semiautomated systems since the rate of scanning should be uniform to interpret the position of flaws from the recorder chart. A, B and C Scans : In the most common flaw detectors the X axis or the horizontal scale represents the distance or depth of flaw and the Y axis or the vertical scale represents the amount of energy reflected. This type of presentation on CRT is called an A-SCAN PRESENTATION. A -scan gives information about defects below a point and we get the depth of the defect and amount of energy reflected. In another type of (special) equipment the probe is moved along a line and the xaxis of the CRT represents the probe position along this line. The vertical scale represents the depth and whenever a defect indication is obtained the CRT trace becomes brighter. In

47

effect the picture gives a cross section (or sectional) view of the object under a line on the surface. Such presentation on CRT is called B-SCAN PRESENTATION. B-scan gives information about the depth of the defect and the extent of the direction in the direction of motion. No information about the amount of energy reflected is obtained. In the third type of presentation, the probe is moved over a surface in a systematic manner. The electron beam also moves over the CRT screen in an identical manner. Whenever a defect indication is obtained the trace becomes brighter. Thus a plan view (or top view) of the test object is obtained. Such presentation is called C-SCAN PRESENTATION. C-scan gives information about the size of the defect in two dimensions but the information about the depth and energy reflected is not obtained. Immersion testing : When large number of components or components of small size are to be tested it is advantageous to use immersion scanning. In this case, the component is immersed in water and the water acts as the couplant. The probe or the workpiece can be moved in a systematic manner and this can be automated. The component in which the probe is held is called the probe holder or probe manipulator. By means of the manipulator the probe angle can be varied. Varying the probe angle in immersion testing is called angulation. The probe along with the manipulator is mounted on an assembly called the 'bridge' which moves in the X and Y direction carrying the probe with it. The bridge is also called the 'scanner'. The immersion scan probe is hermetically sealed to prevent water from entering inside. Immersion probes are designed to work with water in front of them and therefore should not be used as contact probes.

48

Chapter-18

DEFECTS ENCOUNTERED IN MATERIALS

Discontinuity Any break or interruption in the normal physical structure of an article is called discontinuity or defect. 1. Casting Defects : a. Porosity - It is round or nearly round, and is caused by entrapped gas in the molten material. Non-metallic inclusions - These are of irregular shape and consist of slate like impurities accidentally included in the molten metal. Pipe - This is caused by shrinkage as the molten metal solidifies. It may extend deeply into the centre of the ingot. Cold-Shut - Cold shut is formed when molten metal meet with already solidified or relatively cold metal. Cold shut can also be formed by the lack of fusion between two intercepting surfaces of molten material of different temperatures. Hot-Tears - A hot-tear is caused by unequal shrinking of light and heavy sections of a casting as the metal cools. Micro-shrinkage - This defect results from contraction during solidification where there is not an adequate opportunity to supply filler material to compensate for shrinkage. Shrinkage cavity - It is caused by a void left in cast metals as a result of solidification shrinkage. It can be found any where in the cast product. Blow holes - Blow holes are small holes on the surface of the casting caused by gas which is not within the molten metal external gas. This external gas comes from the mold itself.

b.

c.

d.

e.

f.

g.

h.

2.

Rolling defects : a. Laminations - Defects with separation or weakness generally aligned parallel to the worked surface of the metal. It may be the result of pipe, blisters, seams, inclusions or segregations elongated and made directional by working. Lamination defects may also occur in metal powder compacts. Stringers - It is the longer and thinner configuration of non-metallic
49

b.

inclusions aligned in the direction of working commonly, the term is associated with elongated oxide or sulfide inclusions in metals. c. Seams - It is unwelded fold or lap which appears as a crack, usually resulting from a defect obtained in casting or in working. Seams are always open to the surface.

3.

Forging Defects : a. Forging Lap - A forging lap is a discontinuity caused by folding of metal in a thin plate on the surface of the forging. It is due to the mis-matching of the mating surfaces of the two forging dies or abrupt changes in grain direction. It is always open to the surface. Forging bursts or cracks - It is a rupture caused by forging at improper temperatures. Forging a metal at a too low temperature may cause this defect. They are either internal or may occur at the surface.

b.

4.

Welded pipe and Tubing discontinuities : a. Seams - It results from lack of fusion at the weld. This seam may appear either on the inside or outside of the pipe. This comes from faulty welding. Laminations - This may be in the plate stock or sloop from which the pipe is made.

b.

5.

Seamless Pipe and Tube discontinuities a. Slug - This discontinuity forms on the inside of the pipe caused by the piercing mandrel. Slug has been deposited in the pipe wall as the result of severe metal build up on the mandrel. Gonging - Caused by friction between the sizing mandrel and the inside surface of the pipe or tube. Seams or stringers - Found in the bar - stock from which the pipe is made.

b.

c. 6.

Extrusion discontinuities : This can result from the extruding process itself. If the metal does not flow through the die properly, there can be cracks or galling in the finished part. This can be the case in hot extrusions as well as in cold extrusions also. Again, if the original bar stock has contained a crack or porosity, the same discontinuities would show up in the formed part also.

7.

Processing discontinuity :
50

a.

Grinding cracks - Grinding cracks can be caused by stresses which are built up from excess heat created between the grinding wheel and the metal. It occur at a right angle (crosswise) to the rotation of the grinding wheel. Cracks have no relation to the grain direction of metal. Grinding cracks may or may not occur across the grain in hard metals. Heat-treating cracks - Unequal cooling between light and heavy sections of a part which is being heat-treated results in cracking. Explosive forming cracks - It is caused by the explosive force overstressed the material while shaping it to the contour of the dia. A discontinuity is most likely to develop at a point in the dia where the most extreme deforming of the stock occurs or where the die changes contour abruptly. Fatigue cracks - These cracks usually start from stress-concentration points which themselves are open to the surface in most of the cases. Nicks, grinding cracks, forging laps, even poorly finished surfaces, are all examples of discontinuities which might result in fatigue cracks. Some fatigue cracks are sub-surfaces which are originating from porosity or non-metallic inclusions.

b.

c.

d.

8.

Welding defects : a. Crater cracks - Crater cracks are caused at the weld bead by improper use of the heat source either when a weld is started or stopped. A crater crack can also occur at the temporary stop of the weld also. There are 3 types of crater cracks ; Star crater crack - A roughly star - shaped crack, Longitudinal crater crack - Crater crack that parallels the direction of the weld-bead. Transverse crater crack - Crater crack that runs across the weld and is limited to the area of the crater. Stress cracks - Stress cracks in welds are the result of stresses created during the cooling of a restrained (rigid) structure. Stress crack usually occurs transverse (across) the weld in a single pass weld and longitudinal in a multiple pass weld. Porosity - It is caused by the gas which remain entrapped in the weld. Slag inclusion - It can occur during arc welding. As the electrode melts so does its oxide coating and it mixes with the molten metal. This result in the weld-bead is slag inclusions.
51

(i) (ii)

(ii)

b.

c. d.

e.

Tungsten inclusions - Excessive current during tungsten arc welding can cause the tungsten electrode to melt and deposit in the weld. This called tungsten inclusions. Lack of penetration - It results from incomplete penetration into the parent metal or may be caused by backing, or by the molten puddle. It occurs at the root of the weld. Lack of fusion - It is a failure of the weld to fuse with the parent metal or may be caused by a failure of the weld passes themselves to fuse. This particular discontinuity occurs farther up in the weld than the root. Undercut - This discontinuity occurs where the welder has melted and flushed out some of the parent metal in the line of fusion. It has to be open to the surface and can be inspected visually. If there is undercut, some of the parent metal has been melted away.

f.

g.

h.

52

TABLE-I LIST OF USEFUL FORMULAE

1. 2. Velocity = frequency x wavelength; Acoustic Impedance Where = Density of the material

3.

(a) Longitudinal wave velocity

(b) Transverse wave velocity VT =

E 1 P 2(1 + M)

G P

(c) Surface wave velocity VS = Where E = young's modulus G = shear modulus M = poisson's ratio

0.87 + 112 . M VT 1+ M Z VE V= = f (1 M ) VL = P (1 + M) (1 2 M )

4.

Z2 Z1 Reflection coefficient R = Z2 + Z1

Where Z1 = Acoustic Impedance of medium 1 Z2 = Acoustic Impedance of medium 2

5.

Transmission Coefficient T = 1 - R =

4 Z2 Z1 ( Z1 + Z1) 2

6.

Snell's law ;

Sin1 V1 = Sin2 V2
2 = Refracted Angle

Where 1 = Incident Angle;

V1 = Velocity of Incident wave in Incident medium V2 = Velocity of Refracted wave in Refracted medium.
53

7.

Beam spread

= Sin-1

Where D = Dia of the probe crystal and K = 0.51 for 50% (-6 dB) reduction in intensity = 0.87 for 90% (-20 dB) reduction in intensity = 1.2 for extreme edge of the beam.

8.

Near zone For resonance : Thickness of material; T = n. /2

9.

10.

dB = 20 log10 Where A2/A1 = ratio of amplitudes.

2 K 2 2 2 A = D A1 4 4

2 <<

54

TABLE-II

DECIBELS
dB 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 % 100.00 112.20 125.89 141.25 158.49 177.83 199.53 223.87 251.19 281.84 316.23 354.81 398.11 446.68 501.19 562.34 630.96 707.95 794.33 891.25 1000.00 dB 0 -1.0 -2.0 -3.0 -4.0 -5.0 -6.0 -7.0 -8.0 -9.0 -10.0 -11.0 -12.0 -13.0 -14.0 -15.0 -16.0 -17.0 -18.0 -19.0 -20.0 % 100.00 89.13 79.43 70.79 63.10 56.23 50.12 44.67 39.81 35.48 31.62 28.18 25.12 22.39 19.95 17.78 15.85 14.13 12.59 11.22 10.00

55

TABLE-III

ACOUSTIC PROPERTIES OF SOME MATERIALS

Material Density (kg/m2) 2 2700 3600 1200 1400 8100 9800 8600 8900 6900 2000 5700 3600 1300 19300 7200 8400 1200 1820 8860 11400 2060 1740 Acoustic Velocities Longitudinal Transverse (m/S) (m/S) 3 4 6320 9000 2500 2590 4430 2180 2780 4700 3500 4600 4400 4260 1920 3240 4600 4760 2300 12800 3530 2160 4720 5790
56

1 Aluminium Aluminium Oxide Araldite Bakelite Brass Bismuth Cadmium Cooper Cast Iron Concrete Barium Titanate Glass Glycerine Gold Gray-Cast Iron German Silver Hard-rubber Beryllium Bronze(5% P) Lead Lithium Sulphate Magnesium

Acoustic Impedance =Density X Longitudinal Velocity (106 kg/m2S) 5 17.0 32.4 3.0 3.6 35.8 21.3 23.9 41.8 24.1 9.2 15.3 2.4 62.5 33.1 40 2.7 23.3 31.2 24.6 9.7 9.8

3130 5500 2120 1100 1500 2260 2200 25.2 2560 1200 2650 2160 8710 2230 700 3100

Mercury Molybdenum Nickel Inconel Monel Plexi-glass (Perspex)

13000 10900 8800 8250 8830 1180

1420 6290 5630 7820 6020 2730 1400 5600 3960 2340 2620 2380 5920 5660 7390 8100 5460 3200 3320 4170 4620 5760 5570 3600 1480 1740 333

3350 2960 3020 2720 1430 2395 3500 1670 925 1080 1150 3250 3120 2990 3120 2620 1670 2410 2360 3515 1590

18.5 63.5 49.5 64.5 53.1 3.2 1060 3.3 13.4 84.7 2.2 2.8 2.5 46.5 45.5 56.7 27.7 104.3 59.8 24.2 29.6 30.0 15.3 14.5 37.8 1.48 1.51 0

Polyvinylchloride (PVC) Porcelain Platinum Polyethylene Polyamid(nylon) Polystyrene 2400 21400 940 1100 1060

Steel (reference block) 7850 Stainless Steel 302 Stainless Steel 410 Titanium Tungsten Uranium Tin Zinc Zircaloy -2 Quartz Quartz glass Silver Water Motor oil Air 8030 7670 4540 19100 18700 7300 7100 6500 2650 2600 10500 1000 870 1

57

Table-IV

USEFUL TRIGONOMETRICAL RATIOS

300 350 450 600 700 800

Sin 0.50 0.57 0.71 0.87 0.94 0.98

Cos 0.87 0.82 0.71 0.5 0.34 0.17

1 Cos 1.15 1.21 1.41 2.00 2.92 5.76

tan 0.58 0.70 1.00 1.73 2.75 5.67

58

Transmitter

Receiver

TRANSMISSION TECHNIQUE

Fig. 1

T+R

PULSE ECHO TECHNIQUE

Fig. 1

59

Fig. 3

Distance

WAVE Fig. 3-A

Direction of P ropogation

LONGITUDINAL WAVE

Fig. 3-B

TRANSVERSE WAVE OR SHEAR WAVE

Fig. 3-C

Particle Oscillation

at root surface

RAYLEIGH WAVE OR SURFACE WAVE

60

Fig. 4

CATHODE RAY TUBE

Fig. 4-A VOLTAGE

TIME

SAW - TOOTH VOLTAGE

Fig. 4-B

THE FLAW DETECTOR - BLOCK DIAGRAM

61

Fig. 5 RANGE CONTROL : LONG RISE TIME

SHORT RISE TIME

Fig. 5-A

DELAY CONTROL UNDELAYED SWEEP

PULSER

DELAYED SWEEP DELAY DELAY

PULSE ENERGY CONTROL :-

Fig. 5-B
LOW ENERGY PULSE (NARROW) HIGH ENERGY PULSE (WIDE)

DAMPLING CONTROL :-

Fig. 5-C

UNDAMPED

DAMPED

62

Fig. 6

Wave form Display control

RECTIFIED UNRECTIFIED (OR RP)

Fig. 6-A

SUPPRESSION CONTROL Suppressed Un-suppressed

Fig. 6-B

FILTERING CONTROL

Un-filtered

Heavily Filtered

Time Corrected Gain TCG not Applied

TCG Applied

Gate and Thresold Shown by Separate Trace AA Fig. 6-C

63

MAKER RANGE GATE

SWEEP DELAY TCG CLOCK INTERFACE PULSER PULSE ENERGY Damping CRT

PRF

AMPL. MONITOR THRESHOLD SUPPRESSION Band Width/Frequency FILTER WAVE FORM GAIN

UT FLAW DETECTOR CONTROLS

Fig. 7

64

GATEWAY

HORIZONTAL LINEARITY :-

0 1 2 3 4 5 0 10 20 30 40 50 mm SCALE IS LINEAR VERTICAL LINEARITY :-

0 1 2 3 4 5 0 10 20 30 40 50 mm SCALE IS NON-LINEAR SOLID LINE-LINEAR SWEEP DOTTED LINE-NON-LINEAR SWEEP

CRT PULSER
PROBE

AMPLIFIER 60%

80% 40% 60%

30%

AMPLIFIER IS LINEAR

15% DEPTH RESOLUTION :-

AMPLIFIER IS NOT LINEAR

GOOD RESOLUTION LATERAL RESOLUTION :A B C A B C GOOD

POOR RESOLUTION

A A B C B C BAD

65

Fig. 8

PROBE POSITIONS THE I.I.W. CALIBRATION BLOCK (THE V-1 BLOCK)

66

MARK AGAINST THIS ON PROBE

100R

I.I.W. BLOCK

DISPLAY

PROBE AT POSITION WHERE HOLE ECHO IS MAXIMUM 450 600 700

67

V1 BLOCK WITH STRAIGHT GROOVE

(V-2 BLOCK) ATION BLOCK R B LI A C E R TU THE MINIA 0

300

5R

450

60

700

CK LIBRATION BLO MINIATURE CA

RANGE CALIBRATION BY V2 BLOCK

50R 25R

50R

Fig. 9

68

125

ASTM
A = METAL TRAVEL DISTANCE D = HOLE DIAMETER I = IDENTICATION

Fig. 10. ASTM BLOCKS

69

f = 10MH3, Probe " Flat Water Distance = 3" 8 " to Test Hole 64
100

Signal Height (%)

80 60 40 20 0 0
4" 4/6 4" 4" 2/6 3/6 5/6 4" 6/6 4" 7/6

8/6 4"

4"

Fig. C

10

20

30

40

50

60

70

80

RELATIVE AREA UNITS

Distance Area Amplitude Curve.

100

Signal Amplitude (%)

80 60 40 20 0 0 1

f = 10MH3, Probe " Water depth = 3" 2 3 " Test Hole Size 64

Flat

DISTANCE IN INCHES FROM TEST BLOCK TO FLAT BOTTOM HOLE

ASTM Distance Amplitude Curve. Fig. B

70

A=

TYPICAL PROBE POSITION FOR RESOLUTION CHECK

FIG. 11. STEP WEDGE

FIG. 12-A. HALF MOON BLOCK

71

FIG. 12-B. STEPPED HALFMOON BLOCK

FIG. 12-C. HALF MOON BLOCK WITH MANY STEPS

72

INITIAL PULSE

BE 1 (5mm)

BE 2

5 BE 1 (4.5mm) BE 2

10

SHIFT 0.5 DIV.

1.0 DIV.

RANGE 0-100 mm 0 mm 50 mm 90 100 mm

RANGE 50-100 mm 0 50 mm 5 75 mm 8 90 mm 10 100 mm

73

FIG. 13. GENERATION AND RECEPTION OF ULTRASOUND

PIEZO ELECTRIC EFFECT

74

CONNECTOR CRYSTAL

CONTACT SURFACE WEAR PROTECTION

NORMAL - BEAM PROBE

Fig. 14
PROBE THREADED CAP WEAR MEMBRANE

COUPLANT

TYPICAL WEAR MEMBRANE ARRANGEMENT

SOUND TRAP CONNECTOR

INCIDENT ANGLE PROBE

CRYSTAL WEDGE CONTACT SURFACE

JOB REFRACTED ANGLE

ANGLE-BEAM PROBE

Fig. 15.

75

CRYSTAL ASSEMBLY

WEDGE

ASSEMBLED PROBE

REMOVABLE WEDGE TYPE

SKIP AND PATH DISTANCE

76

FLAW LOCATION BY ANGLE BEAM PROBE

Fig. 16

SKIP & PATH DISTANCE IN PIPE

AXIL SCAN OF PIPE

CIRCUMFERENTIAL SCAN

Fig. 17

77

PULSER

AMPLIFIER

DUAL PROBES

ROOF ANGLE

ACOUSTIC BARRIER DEAD ZONE CROSS - TALK

EFFECTIVE RANGE

Fig. 18

78

TRANSDUCER

Fig. 19

INTENSITY

AXIAL DISTANCE

RADIO DISTANCE

INTENSITY

79

NO OVERLAP

NO OVERLAP

1, 2, 3, PROBE POSITIONS

TRANSDUCER

80

Fig. 20 NORMAL INCIDENCE AT AN INTERFACE : REFLECTION AND TRANSMISSION. R = Er/Ei = (Z1 - Z2)2 T = Et/Ei = 4Z1 Z2)2

/ (Z1+ Z2)2

/ (Z1+ Z2)2

OBLIQUE INCIDENCE : REFLECTION AND TRANSMISSION AND MODE CONVERSION

Fig. 21
81

4mm

6mm DGS SCALE

2mm

50

100

150

200

250 mm

SIDE DRILLED HOLE

EFFECT OF SHAPE

Orientation
MAXIMUM ECHO SMALL ECHO

NO ECHO
82

SPRING

OUTLET

PROBE

BUBBLER PROBE SHOE SPRING LOADED PROBES TEST OBJECT WATER INLET

AMPLITUDE

A-SCAN DEPTH

DEPTH

SIDE VIEW B-SCAN PROBE MOVEMENT

DEFECT PROBE POSOTION FOR A-SCAN PROBE MOVEMENT FOR B-SCAN PROBE MOVEMENT FOR C-SCAN
ABOVE MOVEMENT

TOP VIEW C-SCAN PROBE MOVEMENT

83

= T TAN B = HALF - SKIP = T SLC B = BEAM - PATH ( V PATH)

S = 2T TAN B = FULL - SKIP B = 2T SLC B = BEAM - PATH (V PATH)

IF B<

V PATH

S1 = B1 SIN B D = B1 COS B
S ! 2

IF B >

V PATH

S2 = B SIN B D = 2T - B COS B WHERE B = B1 + B2

SELECTION OF ANGLE PROBES Plate Thickness (T) in M.M. 5 to 15 15 to 30 30 to 60 > 60

84

Probe Angle (B) in Degree 800 700 600 450

85