You are on page 1of 21

British Standard

A single copy of this British Standard is licensed to

Tom Magee
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

06 March 2004

This is an uncontrolled copy. Ensure use of the most

current version of this document by searching British
Standards Online at

Methods for

of Welds
Part 2: Automatic examination of fusion
welded butt joints in ferritic steels
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

UDC 621.791.05:620.179.16 + 669.15194.57:621.791.052.4:621.791.5/8:620.179.16

BS 3923-2:1972

Co-operating Organizations

The Welding Industry Standards Committee, under whose supervision this

British Standard was prepared, consists of representatives from the following
Government departments and scientific and industrial organizations:

Aluminium Federation Institution of Civil Engineers

Associated Offices Technical Institution of Electrical Engineers
Committee* Institution of Mechanical
Association of Consulting Engineers
Engineers Institution of Production Engineers
British Constructional Steelwork Institution of Structural Engineers
Association Lloyds Register of Shipping*
British Electrical and Allied London Transport Executive
Manufacturers Association* Ministry of Defence, Combined
British Railways Board Ministry of Defence, Navy
British Steel Industry* Department*
Crown Agents for Oversea Shipbuilders and Repairers
Governments and National Association*
Administrations Society of British Aerospace
Department of Employment Companies Limited*
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

Department of Trade and Industry Welding Institute*

Department of Trade and Industry,
National Engineering Laboratory

The Government department and scientific and industrial organizations

marked with an asterisk in the above list, together with the following, were
directly represented on the committee entrusted with the preparation of this
British Standard:

British National Committee for Non-destructive Testing Society of

Non-destructive Testing Great Britain
Electricity Council, the Central Process Plant Association
Electricity Generating Board Society of Motor Manufacturers and
and the Area Boards in Traders
England and Wales Society of Non-destructive
Engineering Equipment Users Examination
Association United Kingdom Atomic Energy
Institute of Physics and the Physical Authority
Society Water-tube Boilermakers
Ministry of Defence, Army Association
This British Standard, having
been approved by the Welding National Coal Board Individual firm
Industry Standards Committee,
was published under the authority
of the Executive Board on
22nd August, 1972

BSI 01-1999 Amendments issued since publication

First published August 1965 Amd. No. Date of issue Comments
First revision August, 1972

The following BSI references

relate to the work on this
Committee reference WEE/34
Draft for comment 70/26486

ISBN 0 580 07436 6

BS 3923-2:1972


Co-operating organizations Inside front cover
Foreword ii
1 Scope 1
2 Definitions 1
3 Operators 1
4 Equipment 1
5 Surface condition 1
6 Parent metal examination 1
7 Weld examination 3
8 Evaluation of imperfections 4
9 Test plates 4
10 Presentation of results 4
Appendix A Guidance on the determination of probe characteristics 5
Appendix B Notes for guidance on the D.G.S. diagram 5
Appendix C Method for setting sensitivities where maximum
sensitivity is required 12
Figure 1 Relationship between ratio of diameter to
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

thickness and angle of ultrasonic beam 2

Figure 2 Typical D.G.S. diagram for normal probes 8
Figure 3 Typical D.G.S. diagram for angle probe
of 2 MHz, 20 mm 22 mm crystal size 9
Figure 4 Typical D.G.S. diagram for angle probe
of 4 MHz, 8 mm 9 mm crystal size 10
Figure 5 Measurement of shear wave attenuation 12
Figure 6 Measurement of transfer loss 12
Publications referred to Inside back cover

BSI 01-1999 i
BS 3923-2:1972


This British Standard forms part of a series dealing with methods for the
non-destructive testing of welds. It does not state when this particular type of
testing should be employed not does it give standards of acceptance, as both of
these aspects should be covered in the appropriate application standard or be
agreed between the contracting parties.
This standard was first published in 1965 as a result of ultrasonic examination
procedures being agreed between organizations interested in the manufacture
and use of boilers and pressure vessels. Part 2 has now been revised to cover
automatic examination of butt joints in a wide range of shapes and forms of
ferritic steels.
The use of a D.G.S. (Distance, Gain, Size) diagram for obtaining the appropriate
sensitivity setting only has been introduced although it is not the sole method.
Guidance on this method is given in Appendix B.
The revised standard is not associated with any particular type of fabrication, but
has been prepared to cover a wide range of products and as such lays down the
broad principles of automatic ultrasonic examination. It is emphasized that a
satisfactory technique can only be determined after taking into account all the
relevant factors regarding the equipment to be used and the characteristics of the
weld to be examined.
Prior to ultrasonic examination the weld should be visually inspected and any
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

visible flaws recorded. Automatic ultrasonic examination is usually

supplemented by manual examination and both are often used in conjunction
with other testing methods in order to examine completely a welded article or
structure. The use of any non-destructive testing method should always be
considered in relationship to inspection and testing as a whole and the full
benefits of this or any other method often can only be obtained by considering the
results in conjunction with those from other methods.
A British Standard does not purport to include all the necessary provisions of a
contract. Users of British Standards are responsible for their correct application.
Compliance with a British Standard does not of itself confer immunity
from legal obligations.

Summary of pages
This document comprises a front cover, an inside front cover, pages i and ii,
pages 1 to 14, an inside back cover and a back cover.
This standard has been updated (see copyright date) and may have had
amendments incorporated. This will be indicated in the amendment table on
the inside front cover.

ii BSI 01-1999
BS 3923-2:1972

1 Scope 4.3.3 For angle probes, the ultrasonic beam shall

have an angle of refraction of between 35 and 70
This Part of this British Standard deals with
within the material.
methods for the automatic scanning and recording
of imperfections in fusion welded butt joints in 4.4 Coupling medium. The coupling shall be
ferritic steels1). obtained by either contact, gap or immersion
scanning using a liquid or paste medium suitable for
Whereas the procedures apply to all thicknesses of this purpose.
material, the limiting parameters are:
4.5 Continuity of coupling. There shall be an
1) whether the material traversed by the
indication of adequate acoustic coupling for
ultrasonic beam is such that the beam can be
example, by recording a compressional wave
propagated through it sufficiently to achieve the
transmitted through the coupling or by a device
agreed sensitivity; which indicates any failure of the coupling.
2) that with the angles of probe available, the 4.6 Recording. The recording or marking system
ultrasonic beam will reflect from the under
shall clearly indicate imperfections which require
surface of the material in such a way as to cover
further investigation by manual scanning.
the full cross section of the weld. For curved parts
the ability of the beam to reflect from the under Any recorded imperfection shall be clearly located
surface depends on the ratio of the outside from a positional datum so that it can be accurately
diameter to the thickness (see Figure 1). positioned along the weld. It is recommended that
these positional data are located along welds at
NOTE The titles of the British Standards referred to in this
intervals of 250 mm.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

standard are listed on the inside back cover.

2 Definitions 5 Surface condition

For the purposes of this British Standard the The condition of the surface that will be in contact
definitions given in BS 3683-4 apply. with the probe shall be such that a satisfactory
coupling between the probe and the workpiece can
3 Operators be maintained. Spatter and loose scale shall be
If required, the operator shall demonstrate to the
satisfaction of the contracting parties that he is able The surface condition of the weld and adjacent
to meet the requirements of any particular parent metal on each surface shall be such that it
technique adopted. does not adversely influence the weld examination.
NOTE Depending on the profile and surface condition, dressing
of the weld area may be necessary, even when contact is only to
4 Equipment be made with the parent metal.
4.1 Presentation. The examination shall be
carried out by an agreed automatic or 6 Parent metal examination
semi-automatic scan, using a pulse echo technique 6.1 General. Whether or not the parent metal has
with either A or B-scope presentation, and a fully been ultrasonically tested previously, manual
automatic recording or marking system shall be ultrasonic examination shall be made after welding:
1) to locate any flaws, such as laminations, in the
4.2 Test frequency. The equipment shall be material through which the ultrasonic beam will
capable of working at a test frequency within the pass during examination of the weld,
range 1 MHz to 6 MHz. and
4.3 Probes2) 2) to establish the material thickness to enable
4.3.1 The area of each transmitting and/or receiving the actual beam paths to be determined.
crystal shall not exceed 500 mm2. No dimension of If any flaws are found, their influence on the
the crystal face shall exceed 25 mm unless inspection of the weld shall be considered and if
otherwise agreed. possible alternative techniques of scanning shall be
4.3.2 For parent metal examination, zero angle used to ensure complete examination of the weld. If
longitudinal wave probes shall be used. any section cannot be so tested, this fact shall be
included in the report (see 6.6).

For the examination of welds used in the manufacture of pipes and tubes on a continuous basis see BS 3889-1A.
Guidance on the determination of probe characteristics is given in Appendix A.

BSI 01-1999 1
BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

Figure 1 Relationship between ratio of diameter to thickness

and angle of ultrasonic beam
During the preliminary examination the operator 6.4 Time base calibration. The time base shall be
shall assess the attenuation characteristics of the calibrated using either the A2 or A3 block as
material and the influence that the surface specified in BS 2704.
condition will have on the coupling, in order to 6.5 Sensitivity. The method of setting the
determine the practicability of performing an sensitivity shall be agreed between the contracting
effective test (see also 6.5). parties.
6.2 Method. The material shall be examined by the Methods of achieving the gain settings are described
pulse echo technique using a longitudinal wave in Appendices B and C.
probe normal to the surface of the material.
The sensitivity of the flaw detector/probe
(See also BS 4336-1A.)
combination shall be such that a clear signal will be
The parent metal shall be ultrasonically examined obtained from the smallest defect to be detected
over the region covering the entire scanning zone for throughout the scanning distance.
the subsequent weld examination.
The size, type and orientation of the smallest flaw to
6.3 Test frequency. The frequency shall be be detected shall be agreed between the contracting
between one and two times the frequency to be used parties before testing is commenced.
for the weld examination so that the attenuation
characteristics assessed under 6.1 are related to the
subsequent shear wave test.
NOTE On thin material it may be desirable to use a double
crystal probe.

2 BSI 01-1999
BS 3923-2:1972

6.6 Reporting. A report on the parent metal 7.4 Equipment calibration

examination shall be made describing any local 7.4.1 Calibration of time base. For equipment using
heterogeneities (e.g. laminations, surface flaws, A-scope presentation, the time base shall be
areas of high attenuation) which will confuse the calibrated in accordance with the recommendations
interpretation of the weld examination at such of BS 2704 using a type A2 or A3 test block.
locations. If no such heterogeneities are found, a
statement that the parent metal examination was Where B-scope presentation is used, direct time
satisfactorily undertaken shall be made. base calibration may not be possible. In this case a
special test block of the same thickness and
7 Weld examination curvature as the seam to be examined, with four
slots to represent the corners of the weld, shall be
7.1 Method. The examination shall be carried out used to ensure that complete coverage of the weld
with angle probes by an agreed automatic or area is obtained.
semiautomatic scanning system, using the pulse
7.4.2 Position of probe index. This shall be
echo technique with either A or B-scope determined in accordance with the
presentation. A fully automatic recording or recommendations of BS 2704. If a variable angle
marking system shall be used.
probe is used, the upper and lower limits of the
For imperfections predominantly parallel to the probe index shall be determined in a similar manner
direction of welding the weld shall be scanned at to that for a fixed angle probe.
right angles to the weld axis.
7.4.3 Angle of refraction. For a fixed angle probe the
For imperfections predominantly transverse to the angle of refraction shall be measured in accordance
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

direction of welding the weld shall be scanned along with the recommendations of BS 2704.
the weld axis.
If a variable angle probe is used, the upper and
NOTE The automatic system may not facilitate the satisfactory lower limits of the range of angle shall be
detection of transverse imperfections. In such cases it will be
necessary for an alternative automatic or manual method to be determined in a similar manner to that for a fixed
agreed between the contracting parties before testing is angle probe.
7.5 Sensitivity. The method of setting the
7.2 Test frequency. The test frequency used shall sensitivity shall be agreed between the contracting
be compatible with the requirements for sensitivity parties.
in 7.5.
Methods of achieving the gain settings are described
NOTE For guidance, the following frequencies are suitable for
the beam path lengths quoted:
in Appendices B and C.
4 MHz to 5 MHz for up to 200 mm The sensitivity of the flaw detector/probe
2 MHz to 2.5 MHz for up to 400 mm combination shall be such that a clear signal will be
1 MHz to 1.25 MHz for over 300 mm obtained from the smallest defect to be detected
7.3 Scanning. Wherever possible the scanning of throughout the scanning distance.
the weld shall be carried out from both sides of the The size, type and orientation of the smallest flaw to
weld on the same surface. Where the configuration be detected shall be agreed between the contracting
is such that examination from both sides is not parties before testing is commenced.
possible, this fact shall be included in the report
(see Clause 10). With variable angle probes a marked loss in
sensitivity may occur at the higher probe angles and
The full cross section of the weld shall be scanned at therefore the sensitivity shall be determined using
the agreed sensitivity. This may be achieved, for that angle which gives the weakest echo.
example, by the use of
7.6 Recording. If the recording system is of
(1) a variable angle probe; the GO/NO-GO type, the echo height shall be
or (2) a to and fro movement of a fixed angle measured at which
probe; 1) the recorded signal appears,
or (3) some special configuration of probes. and
The distance between successive scans of the weld 2) the recorded signal disappears,
shall not exceed the breadth of the transmitting
crystal. and there shall not be a difference of more
than 2 dB.
The speed of movement of the probe(s) shall be such
that the optimum response of the recording With a proportional type recording system, the
apparatus is obtained. recording indication shall be compared over its full
range, with the echo height on the flaw detector.

BSI 01-1999 3
BS 3923-2:1972

8 Evaluation of imperfections 10 Presentation of results

If an imperfection is revealed by the examination In the report of the automatic ultrasonic test the
described in Clause 7, a more detailed ultrasonic following information shall be included:
examination shall be carried out using an agreed 1) Job identification reference (including contract
manual technique which is suitable for the purpose, number and part number).
such as one of those given in BS 3923-1.
2) Surface condition (as welded, chipped, ground
9 Test plates
3) Date of test.
Due to the restricted width normally required,
welded coupon test plates are not suitable for 4) Identity of operator.
automatic scanning, and their examination shall be 5) Instrument used (including serial number).
carried out by an agreed manual technique. 6) Probes used, including frequency.
7) Scanning techniques.
8) Sensitivity of test. (This shall be recorded at
the end of the test and related instrument
settings shall be compared against and noted
from a suitable test block.)
9) Results obtained.
10) Any other relevant information relating to the
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI


4 BSI 01-1999
BS 3923-2:1972

Appendix A Guidance on the A.2 Method of measurement of near field

determination of probe A.2.1 Method 1 (if an immersion tank is available).
characteristics Plot the distance/amplitude curve of a small
reflector along the axis of the beam. The near field
A.1 Method of measurement of frequency. The length, N, is shown by the maximum amplitude
usual method of measuring-probe frequency is to position before the curve falls away into the far field.
take the first back wall echo from a steel sample A useful target for this operation is a 2 mm or 3 mm
such as the A2 block of BS 2704 and to take this diameter ball, mounted on the end of a rod whose
high frequency pulse from the end of the amplifier diameter is less than the ball at the mounting point.
before rectification and present it on the calibrated The near field length for steel can be calculated from
time base of an oscilloscope. The number of full wave the empirical determination of N for water since the
lengths in a one microsecond time interval gives the ratio of the two near fields is inversely proportional
probe frequency. to the ratio of the velocities, i.e.:
Alternatively, the ultrasonic operator, without any
N ( steel ) Velocity ( water )
electronic knowledge, should be quite capable of -------------------------- = -------------------------------------------
measuring the probe frequency provided that the N ( water ) Velocity ( steel )
flaw detector has a fast time base expansion facility, A.2.2 Method 2 (if an immersion tank is not
such as 10 mm steel full screen, and sufficient available). In this case, the distance/amplitude
resolution on each echo to indicate the high curve of the back echo is plotted from flat parallel
frequency pulses. In this case, the frequency is samples of steel of similar velocity whose width
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

derived from the formula: would not restrict the beam. The curve should be
plotted on logarithmic paper and after
Frequency = -------------------------------- approximately three near field lengths, it should
form a straight line. If not, attenuation has to be
For boiler plate the velocity in compression is allowed for (see B.1.1), and when taken into account
usually in the order of 5.93 103 m/s and the for the back wall echo a straight line should result
velocity in shear is in the order of 3.23 103 m/s. in the far field past three near field lengths. If the
The velocity is known for the material and the straight line is then projected up to the zero decibel
wavelength is read from the trace as the distance N
line, the intersection point will be -------- or 1.57 N,
covered by two consecutive half cycles for a half 2
where N is the near field length.
wave rectified display. For flaw detectors which
show a full wave rectified display, it would be the Appendix B Notes for guidance on the
distance covered by four half cycles. From this it
would appear that two wavelengths were being
D.G.S. diagram
measured but in fact normal time-base calibration
B.1 Normal probes. (See Figure 2.)
for flaw detection purposes is always half of the total
distance travelled; the half calibration therefore B.1.1 General. The D.G.S. diagram for single crystal
cancels out the double wavelength. The half cycles normal probes was drawn by plotting amplitude in
should be regularly spaced, uneven spacing decibels from a series of disc shaped reflectors with
indicating a distorted pulse, and probes exhibiting increasing distance from the probe in water. The
this feature should not be used with the D.G.S. loss due to water attenuation was allowed for in
system. each case and therefore the graph shows the
reflection conditions for any material assuming no
The distance is given in near fields and is on a log
base. If the near field length N is not known for the
probe in use it may be calculated from the formula
N = -----------------------------------------
4 wavelength
where D is the crystal diameter.

BSI 01-1999 5
BS 3923-2:1972

The backwall echo line indicates the maximum The term small reflectors is related to the beam
reflection from a large reflector with increasing width and not to the crystal diameter, since
distance from the probe, which becomes a straight reflectors greater than the probe diameter behave
line after three near field lengths. This confirms the as small reflectors when far enough away to be
radiation law for large reflectors in the far field, i.e. encompassed within the spreading beam width.
the amplitude is inversely proportional to the This is shown in Figure 2 by the lines with S values
distance, so that if the distance is doubled the of 1 and 2 times the crystal diameter.
amplitude is halved: a 6 dB reduction. This law B.1.2 Setting the sensitivity. In order to use Figure 2
provides a simple method of measuring attenuation. to determine the sensitivity setting, the operator
If a difference of more than 6 dB is measured, then first needs to know the smallest disc shaped flaw or
this will be due to attenuation in the material, equivalent reflector size that has to be detected and,
leaving a simple calculation for the attenuation secondly, at what amplitude he wishes to record it.
factor. The following plate testing example illustrates the
Example 1 method:
Plate thickness 30 mm Example 2
Probe frequency 4 MHz Plate thickness 100 mm
Probe diameter 10 mm Probe frequency 4 MHz
Near field length 17 mm Probe diameter 25 mm
The second back echo at 60 mm distance is greater Near field length 104 mm
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

than 3N (51 mm), therefore, the second and fourth Smallest flaw to be detected equivalent to
echoes are on the 6 dB slope. flaw 5 mm diameter (0.2S). Defect echo height 2
Suppose the difference in echo amplitude measured screen divisions.
between second and fourth echo = 10 dB. Total beam The first back echo will be at a distance of
path between second and fourth echo = 120 mm. approximately 1 near field. It can be seen from
Therefore, the attenuation of the plate material Figure 2 that close to the surface an 0.2S flaw will
at 4 MHz is give an echo approximately 26 dB below the back
10 6 1 echo set to full screen height.
---------------- = ------ dB/mm or 33 dB/m
120 30 Near the bottom of the specimen such a flaw will
Alternatively, back echoes in the near field may be only be 17 dB below but at the worst position, 0.5N
used. Any difference between their amplitude in distance (52 mm), it will be 29 dB below.
excess of what is shown in Figure 2 indicates the In order to find the flaw at 2 screen divisions at the
amount of attenuation over the total distance worst point (29 dB below), the back echo has first to
travelled between the two echoes. be set to 2 screen divisions and then increased
NOTE The attenuation factor in the example 1/30 dB/mm by 29 dB. This calibration ensures that all flaws
means that the sound is attenuated by 1/30 dB for every that have to be detected will give a signal of not less
millimetre of sound path travelled, bearing in mind that the than 2 screen divisions height, whether or not there
sound in a reflection technique has to travel there and back.
is attenuation present.
Since the flaw detector calibration is always for the
It is obvious when the minimum record level is
distance in front of the probe and not the total
drawn on the diagram that reflections from certain
distance travelled, it may be more convenient to
depths could give echoes greater than that level and
express the factor in the same way, i.e. 1/15 dB/mm
yet still be acceptable. To determine therefore
or 66 dB/m. It is important to indicate in which form
whether or not the echo is greater or less than the
the factor is being used.
minimum size or in fact to determine its actual
The lines below the back echo line show the equivalent flaw size, the gain should be reduced by
amplitude of reflectors which are smaller than the the decibel difference between the 0.2S line and the
beam width. These values are shown in relation to record level at the defect depth. The 2 screen
the crystal diameter and, for small reflectors within divisions is then the minimum record level at the
the near field, the nodes and anti-nodes of distance.
sensitivity are clearly indicated. They do not arise
with a large reflector since all the incident energy is
totally reflected.
Well into the far field all these lines indicate that
reflections from small reflectors follow the inverse
square law, so that if the distance is doubled then
the amplitude is reduced by a factor of 4 (12 dB).

6 BSI 01-1999
BS 3923-2:1972

This example assumes no attenuation in the B.2 Angle probes. (See Figure 3 and Figure 4.)
material at the frequency in use, although by taking
B.2.1 General. The D.G.S. diagram for angle probes
the back echo as a reference any attenuation
is essentially the same as for normal probes, except
present would only make the sensitivity too high that distance is measured from the probe index and
rather than too low. If the attenuation has been therefore most of the near field is fortunately lost in
measured it may turn out to be insignificant
the probe. The horizontal axis as before gives the
1 - dB/mm (5 dB/m) in view of the thickness
e.g. --------- distance along the beam path; corresponding
200 equivalent surface distances are shown for the
involved, If the attenuation was significant, different angle probes. Note that in the case of the
however, as shown in Example 1 (1/30 dB/mm) then two D.G.S. diagrams shown in Figure 3 and
this should be taken into account as follows: Figure 4, the probe size is stated and therefore the
Draw in the attenuation curve as shown. In this equivalent flat-bottomed hole sizes and the
case, since the back echo has been taken as the horizontal axis have been calibrated directly in
reference reflector, the allowance for attenuation millimetres.
is greater the nearer the reflector is to the In order to use these diagrams to determine the
surface. The difference at the surface in this case sensitivity setting the operator first needs to know
is 6.7 dB i.e. (100 mm 2 1/30). the smallest disc-shaped flaw or equivalent reflector
Assuming now that a reflector in excess of the record size that has to be detected and secondly, at what
level was found at a depth of 0.3N, two corrections amplitude he wishes to record it.
should be made to the gain: 4 dB for the difference To set the sensitivity for weld testing with an angle
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

between the 0.2S line and the record level, and 5 dB probe however, is more complex than using a simple
for the record level to attenuation line, i.e. the total normal probe. Because there is no backwall echo to
correction to be made is a reduction in echo height use as a standard reference, a separate reference
of 9 dB after which the decision as to whether or not echo has to be used, such as the 100 mm radius of
the flaw is rejectable is directly related to the record the A2 calibration block (see BS 2704). This means
level of 2 screen divisions. that an allowance has to be made for the different
If it is necessary to determine the actual equivalent surface conditions (transfer loss) between the test
flaw size for the flaw echo at 0.3N, its amplitude in block and the plate, the different attenuation factors
decibels above the 2 screen divisions should be between the two materials and also the difference in
measured and reduced by the decibel value, record distance between the 100 mm radius reflector and
level to attenuation line. For example: the farthest defect distance in the material.
Flaw depth 62 mm = 0.6N The procedure to arrive at the correct gain setting to
Echo height above 2 screen division = 19 dB ensure that the smallest defect to be detected
throughout the scanning distance will at least give
Record level to attenuation line = 3 dB an echo at the appropriate screen height, taking into
Therefore, final position above record account the factors mentioned above, is as follows:
level = 16 dB (on 0.4S line) 1) Measure attenuation of material (see B.2.2).
From Figure 2, S crystal diameter = equivalent 2) Measure transfer loss between A2 test block
flat-bottomed hole diameter and plate (see B.2.3).
Therefore: 0.4 25 = 10 mm diameter 3) Read difference in decibels between back echo
NOTE 1 In the example used the back echo happened to be from 100 mm radius on A2 block to minimum
at 0.96N but it could have been 0.4N or 5.5N or any other value,
depending on the material thickness and the probe near field
defect size at worst position on D.G.S. diagram
length. Furthermore, the S values and the distance values in which should be available for the probe in use.
near fields could both be given directly in millimetres once the 4) Set back echo from 100 mm radius to 2
probe crystal diameter and near field length are known.
NOTE 2 The D.G.S. diagram shown in Figure 2 is based on the divisions, plus difference in decibels from (3) plus
established sound radiation laws. It is essential, therefore, that transfer loss from (2) plus difference in
the amplifier and probe characteristics in use do not give attenuation from 100 mm radius on A2 block to
conflicting results from those shown on the diagram. In other maximum testing distance in plate.
words, the flaw detector and probe combination should be
entirely compatible with the D.G.S. diagram in use.

BSI 01-1999 7
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
BSI 01-1999 8

Figure 2 Typical D.G.S. diagram for normal probes

BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
9 BSI 01-1999

BS 3923-2:1972

BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
BSI 01-1999

Figure 4 Typical D.G.S. diagram for angle probe of 4 MHz, 8 mm 9 mm crystal size
BS 3923-2:1972

Example 3 3) Measure the difference between through

Probe of 4 MHz, 8 mm 9 mm crystal size transmission signal height A. B1 and A. B2.
Smallest defect to be detected equivalent 4) Subtract the difference in decibels indicated on
to 2 mm diameter disc-shaped flaw the D.G.S. diagram, the remainder being
attenuation over half skip distance on the screen
Maximum defect distance 150 mm
or 100 mm of beam path.
Transfer loss 4 dB
Example 5
Attenuation of A2 block 40 dB per metre beam
path in front of probe Probes of 4 MHz, 8 mm 9 mm crystal size
Attenuation of plate 80 dB per metre beam path 1
in front of probe -- skip distance 50 mm
The setting required to bring a 2 mm diameter Signal height A. B1 44 dB
equivalent flaw to the record level at the worst
Signal height A. B2 54 dB
distance (150 mm) would be an echo from ------
the A2 block 100 mm radius to 2 divisions, 10 dB
plus 30 dB from (3) plus 4 dB from (2), plus 8 dB Difference according to D.G.S.
difference in attenuation from 100 mm radius on A2 diagram, 5 dB (50 mm to 100 mm)
block to maximum testing distance in plate,
i.e. an A2 back echo to 2 divisions plus 42 dB.
10 dB 5 dB
With this sensitivity, the minimum equivalent flaw attenuation = ----------------------------------- = 1 10 dB mm of
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

50 mm
size to be recorded will at least reach 2 screen
divisions, wherever it appears throughout the beam path on screen.
scanning distance. It is possible as with the normal B.2.3 Measurement of transfer loss. The transfer
probe test that echoes in excess of 2 screen divisions loss between the A2 block and the plate under test
could be acceptable. To determine whether or not can be measured as follows:
this is so, and at the same time to evaluate the
1) Use one probe of the same angle and frequency
equivalent flaw size for any such echo, proceed as
as will be used for the test and calibrate the
screen for beam path.
a) Draw on the diagram the attenuation curve as
2) Use two probes, separate transmitter and
receiver, and locate as shown in Figure 6.
b) Locate the defect depth on the diagram in
3) Measure difference between through
distance and amplitude above the record level.
transmission signal height of A2 block and plate.
c) Reduce the defect position by the difference in
4) Subtract the difference indicated on the D.G.S.
decibels between the record level and the
diagram, the remainder being transfer loss.
attenuation line.
Example 6
Example 4
Defect echo 50 mm beam path from screen Probes of 4 MHz, 8 mm 9 mm crystal size
Echo height 34 dB greater than record level Through transmission signal
of 2 screen divisions
( 1-- skip on screen), A2 block 56 dB
Record level to attenuation line 8 dB 2
Therefore, new position on Through transmission signal
diagram = 34 8 = 26 dB above record level and
hence equivalent flaw size is 4 mm diameter. ( 1-- skip on screen), plate 68 dB
12 dB
B.2.2 Measurement of shear wave attenuation. The
shear wave attenuation of the plate can be Difference according to D.G.S.
measured as follows: diagram, 3 dB (50 mm to 80 mm)
1) Use one probe of the same angle and frequency Attenuation difference according to attenuation
as will be used for the test and calibrate the curve for A2 block at 50 mm and attenuation
screen for beam path. curve for plate at 80 mm, 5 dB.
2) Use two probes, separate transmitter and Therefore, transfer loss = 12 3 5 = 4 dB.
receiver, and locate as shown in Figure 5 so that
signal A. B1 appears at 50 mm and signal A. B2
at 100 mm on the screen.

BSI 01-1999 11
BS 3923-2:1972

Appendix C Method for setting 2) The probe is next placed on the A2 calibration
sensitivities where maximum block (see BS 2704) and a record made of the
difference in the attenuator reading required to
sensitivity is required bring the echo from the 100 mm radius for angle
After following the requirements of the standard for probes, or the 100 mm parallel section for normal
a particular probe selection and time-base probes, down to full screen height.
calibration the sensitivity settings should be made 3) The attenuator difference between the 100 mm
as follows: radius for angle probes, or the 100 mm parallel
1) The probe is applied to the test surface and the section for normal probes, of the A2 block and
gain controls are adjusted so that the grass that required to raise an echo from a 1.5 mm
level of the material grain structure is drilled hole at an appropriate beam path
raised 2 mm high above the base-line at the distance, is recorded.
appropriate testing distance. Any reduction or It should be stated whether this drilled hole is
increase in the sensitivity to facilitate the parallel to or at right angles to the test surface.
exploration of defects is to be noted in decibels.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

Figure 5 Measurement of shear wave attenuation

Figure 6 Measurement of transfer loss

12 BSI 01-1999
BS 3923-2:1972

Example of maximum sensitivity record

Calibration: Screen calibrated to 250 mm full

path length
Probe selection: 2 MHz, 45, single crystal
Testing To give grass level up
sensitivity: to 250 mm path
A2 calibration (100 mm radius) echo
block: requires 28 dB inserted to bring
it down to full screen height
Test block with Hole was drilled at 125 mm path
1.5 mm hole: length parallel to the test
surface; requires 16 dB inserted
from testing sensitivity level to
give full screen height from this
Allowance can then be made for the transfer loss
and the average attenuation of the test piece as
described in Appendix B.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

BSI 01-1999 13
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

BS 3923-2:1972

BSI publications referred to in this standard:

This standard makes reference to the following British Standards:

BS 2704, Calibration blocks and recommendations for their use in ultrasonic flaw detection.
BS 3683, Glossary of terms used in non-destructive testing. Part 4. Ultrasonic flaw detection.
BS 3889, Methods for non-destructive testing of pipes and tubes. Part 1A. Ultrasonic testing of ferrous pipes
(excluding cast).
BS 3923, Methods for ultrasonic examination of welds. Part 1. Manual examination of fusion welded butt
joints in ferritic steels.
BS 4336, Methods for non-destructive testing of plate material. Part 1A. Ultrasonic detection of laminar
imperfections in ferrous wrought plate.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

BSI 01-1999
BSI British Standards Institution
| BSI is the independent national body responsible for preparing British Standards. It
| presents the UK view on standards in Europe and at the international level. It is
| incorporated by Royal Charter.
| Revisions
| British Standards are updated by amendment or revision. Users of British Standards
| should make sure that they possess the latest amendments or editions.
| It is the constant aim of BSI to improve the quality of our products and services. We
| would be grateful if anyone finding an inaccuracy or ambiguity while using this
| British Standard would inform the Secretary of the technical committee responsible,
| the identity of which can be found on the inside front cover. Tel: 020 8996 9000.
| Fax: 020 8996 7400.
| BSI offers members an individual updating service called PLUS which ensures that
| subscribers automatically receive the latest editions of standards.
| Buying standards
| Orders for all BSI, international and foreign standards publications should be
| addressed to Customer Services. Tel: 020 8996 9001. Fax: 020 8996 7001.
| In response to orders for international standards, it is BSI policy to supply the BSI
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

| implementation of those that have been published as British Standards, unless

| otherwise requested.
| Information on standards
| BSI provides a wide range of information on national, European and international
| standards through its Library and its Technical Help to Exporters Service. Various
| BSI electronic information services are also available which give details on all its
| products and services. Contact the Information Centre. Tel: 020 8996 7111.
| Fax: 020 8996 7048.
| Subscribing members of BSI are kept up to date with standards developments and
| receive substantial discounts on the purchase price of standards. For details of
| these and other benefits contact Membership Administration. Tel: 020 8996 7002.
| Fax: 020 8996 7001.
| Copyright
| Copyright subsists in all BSI publications. BSI also holds the copyright, in the UK, of
| the publications of the international standardization bodies. Except as permitted
| under the Copyright, Designs and Patents Act 1988 no extract may be reproduced,
| stored in a retrieval system or transmitted in any form or by any means electronic,
| photocopying, recording or otherwise without prior written permission from BSI.
| This does not preclude the free use, in the course of implementing the standard, of
| necessary details such as symbols, and size, type or grade designations. If these
| details are to be used for any other purpose than implementation then the prior
| written permission of BSI must be obtained.
| If permission is granted, the terms may include royalty payments or a licensing
| agreement. Details and advice can be obtained from the Copyright Manager.
| Tel: 020 8996 7070.
389 Chiswick High Road |
London |
W4 4AL |