NDTnet - December 1996, Vol.1 No.12
Characterization of Reflectors by Ultrasonic Methods
Here is a comprehensive review of the methods being
used at present or under development, for qualitative interpretation of
reflectors by means of pulse-echo techniques.
Ultrasonic inspection is used to determine whether or not
a test piece may be used according to its intended purpose
("fit for purpose"); that is, whether it is free of
discontinuities or contains defects.
The definition of "defects" cannot be determined solely
from ultrasonic information : knowledge of the construction of
the test piece, its intended purpose, its material and its
fabrication process is necessary. All this information is
important to decide whether an inspected part may be used
Therefore, in the following text the term "defect evaluation"
will not be used if we are referring to the interpretation of
ultrasonic pulse echoes.
2. The pulse-echo method - general remarks
Using the pulse echo method, in principle, all information
about the reflector can be derived only from the echo signal.
Simple evaluation methods use only parts of the resulting information.
According to fig. 1, the following will be available:
- the position of the transducer, when an echo occurs;
- the directional characteristics of the transducer;
- the transit time of the pulse echo;
- information based on the shape of the echo.
From the interaction between the sound waves and the reflector,
three spheres of influence can be differentiated (1,2):
- The transducer influences the transmitted and received pulse.
It has a directional characteristic;
- The reflector influences the shape and direction of
propagation of the reflected pulse, and could even cause
- The material also influences the shape and amplitude of
the echo pulse by sound absorption, anisotropy and scattering
(and sometimes also the direction of propagation).
The interaction between the sound wave and the reflector does not always take place
in an ideal manner as shown in fig. 1 (above). In most cases a portion of the sound beam
strikes the reflector and sometimes only part of the reflected
waves can be received. It is also possible that mode conversion
has taken place and the wave type received is different from
the transmitted pulse. Therefore, when using simple evaluation methods,
the many possibilities of interaction sometimes lead to an
uncertain or even wrong result.
However, the opposite can also be true : if the type and position of a possible
reflector is known, then these simple evaluation methods can give just as good results
as the more sophisticated analytical methods.
3. Simple evaluation methods
3.1. Evaluation of the transit time:
With crack growth in tensile test pieces, the reflector type
and position is known. The crack depth can be determined from
the measurement of the transit time (3,4) (see fig. 2).
Stress cracks on parts having simple shapes can be determined
in a similar way if the direction of crack propagation is known.
3.2. Reflector edge scanning:
If information about transducer position is added
to the transit time data, then one obtains a scanning method
with which the dimensions of large reflectors can be determined.
This method is used when the size of the detected reflector is
distinctly larger than the sound beam diameter (fig. 3).
Fig. 2 - Simple transit time method (End-on-crack)
Fig. 3 - Simple reflector scanning (amplitude drop by x dB)
From the transit time, the depth position of the reflector can be derived,
the measurement of the transducer scanning track will give reflector
expansion. In most cases the projection of the reflector area
on the surface of the test piece is used (C-scan method).
However, the echo amplitude must also be considered in order to determine
the reflector edge. It is assumed that the reflector edge
is positioned under the centre of the transducer when the echo
amplitude has dropped below its maximum by a predetermined value of x dB
(5,6,7), i.e. by 6 dB (the half value method).
An amplitude decrease of 20 dB is also quite common (8).
3.3. Evaluation of the maximum echo amplitude:
If the reflector area, contrary to paragraph 3.2., is smaller than
the sound beam diameter, then evaluation is accomplished using
the maximum echo amplitude combined with appropriate transit time information.
This is the most common evaluation method used today in manual testing.
The transit time determines the reflector position and the maximum
echo amplitude determines a (fictitious) reflector size. The DGS-method
uses the ideal circular reflector as an equivalent reflector.
Independent of the actual reflector type and possible inclined position,
the echo is evaluated as if it had come from a circular reflector of
equivalent size which was hit perpendicularly (9).
In addition to the uncertainties which the use of an equivalent
reflector imply, the rectified video signal is normally used rather
than the RF-echo presentation (fig. 4). Properties of the electrical
transmission line also influence the result (rectification, smoothing,
Fig. 4 - Simple evaluation of echo amplitude (DGS method)
3.4. Reflector scanning using automatic methods :
Information about probe position, time of flight and
maximum echo amplitude can be automatically obtained
by scanning the test object, storing the results and
evaluating them with a computer (10,11).
COMSON is a system which carries out a reflector diagnosis
by which the echo amplitude is evaluated over different sound
paths when testing in the manual mode (12). Fig. 5 shows an example of this.
Fig 5: - Multiple-path scanning by the COMSON method. (44Kb)
Rapid beam steering (parallel to surface movement
and swivelling) is carried out electronically using
phased arrays (13, 14, 15).
The reflector size is mostly displayed as a plane area. Fig. 6 shows the principal method :
- Linear scanning by shifting the sound beam in a line or within a plane;
- Sector scanning by swivelling the sound beam within a certain angle range;
- Compound scanning by combining linear scanning with sector scanning.
Fig. 6 - Scanning a reflector by phased arrays
The same reflector can be detected a number of times from different locations.
If the location of a reflector changes within the sound beam,
then characteristic changes occur with echo transit time and amplitude.
The ALOK method (Amplitude -Time of Flight- Local Curve Method) makes
use of this physical law (16, 17, 18), (refer to fig. 7).
|| Fig. 7 - Evaluation by the ALOK-method |
Defect border reconstruction or
or pattern recognizition (classification)
- transmitter/receiver characteristics
The amplitude value, A=f(s), and the time of flight, L=f(s),
received via the probe shifts, are freed from any interference by
using a computer. The influence from the receiver and transmitter probes
is eliminated. A mathematical reconstruction is made
of the reflector edge or characteristic values which one can process using a
pattern recognition program. Reflector classification is then automatically
carried out. This also applies to the COMSON method.
3.5. Multiple frequency method
There is another analysis possibility by which the ultrasonic pulse changes
its spectral composition, because reflection behaviours of plane and
inclined reflectors are strongly dependent on frequency.
An analysis of the echo signal, with dependence on the frequency, can also
give reliable results in manual testing especially when the reflectors do not
have ideal reflective characteristics, e.g. castings or welds (19, 20).
Fig. 8 shows reflection behaviours at different frequencies (according to (19)):
- At very high frequencies, only small areas of the fissured reflector return to the probe (shaded).
- At low frequencies, these areas become larger.
- At very low frequencies the complete area of the reflector is reflecting back to the probe (transition from reflection to scattering).
4. Evaluation of diffracted and converted sound waves
The considerations made up until now have assumed that a
plane wave hits the reflector which then reflects back only
this wave mode in an ideal sort of way (Fig. 9a).
Fig. 9a - Taking into account of diffraction phenomena
Fig. 9b - Taking into account of diffraction phenomena
In reality, with reflections from real reflectors, diffracted waves
appear, mostly from the edge of the reflector. Depending on the
wave's angle of incidence onto the reflector surface, other types
of waves can be generated : longitudinal or transverse
waves, even surface waves. This is illustrated in Fig. 9b.
Basically, information about the reflector can be obtained from
all waves produced, whether it be from the times of flight or from
the echo amplitudes.
Using focused sound fields, diffracted waves can be greatly excited at the edge of the reflector (Fig. 10).
By scanning the edge, the reflector size and angle can
be determined from the time of flight and the amplitude (when
the reflector is larger than the diameter of the sound beam)(21).
Fig. 10 - (Focal distance / diameter)
Fig. 11 - Diffraction at crack tips time-of-flight technique
In spite of the term "superintensity", it is the sound pressure
amplitude which is evaluated and not the intensity of the sound.
Even if it is only possible to evaluate the times of flight from the diffracted sound waves, an accurate measurement can be made of the reflector dimension.
This is of special interest regarding natural cracks where there is
no clear relationship between echo amplitude and crack extension.
This type of evaluation, from the time of flight, is now
known as TOFD (Time-Of-Flight Diffraction technique) (22, 23).
Figure 11 shows detection of the crack tips by determination of
the diffracted wave's time of flight to various probes or probe
positions for surface cracks and for internal cracks as well.
An overview of possible ultrasonic methods for determination of
the crack extension is explained in (24).
|In order to separate diffracted sound pulses from the tips of smaller reflectors (regarding time of flight) short, broad band pulses must be used (25).
The phase information from a RF pulse, which is diffracted at the tips
of the reflector, can determine whether it is a connected reflector,
a flat reflector, or a voluminous reflector (26).
This determination is especially important in the field of weld testing.
Figure 12 shows the decision tree using this method of evaluation.
Fig. 12 - Evaluation of diffracted waves (Radio frequency pulse)/Sign of first oscillation
It should be noted here that the COMSON method (12) uses rectified pulses.
If incoming waves are converted to surface waves at the reflector,
as seen in Fig. 13a, then these will produce additional echoes
(longitudinal and transverse waves) from the reflector
boundaries, which will return to the probe later than direct echoes
from the edge of the reflector. This is the reason why they are
called "satellite echoes" (Fig. 13b).
With cracks, the surface wave propagates along the crack surface;
with spherical or cylindrical reflectors it travels around them (27, 28)
Fig. 13a Fig. 13b - Mode conversion satellite pulses
From the type of satellite echo and its time of flight, the type and
extension of a reflector can be determined, e.g. differentiation
between pores and cracks (29-31). Converted waves are reflected
at another angle than the original incoming waves. By application
of crystals for two wave modes (longitudinal and transverse waves),
which are arranged at various angles, the dimensions of reflectors need
only be measured using one probe (containing two crystals).
This is of advantage with reflectors which are not positioned
perpendicularly to the striking wave and which, up until
now, could only be detected using the tandem method (with
two probes). Due to longitudinal-longitudinal-transverse
mode conversion (Fig. 14), people speak of the LLT technique (32)
or, according to the type of probe, of multiple-beam-multiple-mode probes (33)
Fig. 14 - Mode conversion with LLT-method
- Serabian, St.: An assessment of the detection ability of the angle beam
interrogation method. - Mater. Eval. 39 (1981), 1243 - 1249
- Wüstenberg, H.; Erhard, A.: Development of ultrasonic techniques
for sizing defects. In: Nichols, R.W.(Ed.): Nondestructive Examination
Relation Structural Integrity.
London, Appl. Sc. Publ., 1980, 59 - 83
- Winter,D.C.: End-on-crack measurement. In: McAvoy, B.R.(Ed.): Ultrasonics
Symposium Proceedings 1975. New York, Inst. Electr. Eng. 1975, 572 - 574
- Hess,A., Thoma,Ch.: Moglichkeiten der Rißtiefenbestimmung mit Ultra-
schall nach dem Rißspitzenverfahren. Materialprüfung 25 (1983), 1O, 340
- Schlengermann,U., Frielinghaus,R.:
Beitrag zur FehlergroBenbestimmung mit Ultraschall durch Fehlerabtastung
mit relativer Schwelle. Materialprufung 15 (1973), 2, 50 - 56
- Schlengermann,U.: Beitrag zur Fehlergroßenbestimmung durch
Fehlerabtastung mit fester SchwelIe. Materialprüfung 16 (1974), 10, 319
- Schlengermann, U.; Frielinghaus, R.: Remarks on the practice of determining
the size of reflectors by scanning. In: ComFranc Etud. Ess. Non Destruct. (Ed.):
Proceedings 8th WCNDT Cannes (1976), report 5 C 10
- Jackson, B.: The use of edge-of-beam methods for the assessment
of defect size. Can. Soc. Nondestructive Testing Journ. 5 (1983),1; 48,50
- Krautkrämer, J.; Krautkrämer, H.: Werkstoffprüfung mit Ultraschall.
Springer Verlag Berlin,1986, 101 -112.
- Anderson, E.B.: The evaluation of defects with P-scan equipment
and focused search units. In: Conf. Period. Inspect. Press.
Components, London, Inst. Mech. Engin., 1982, 163-168.
- Rensmeyer, M.E.; Grothues, H.L.: Automated data processing for
ASME section Xl requirements using the SUTAR system.
In: ASNTFall Conf. Proceed., Denver, Am. Soc. Nondestr. Test., 1978, 175-181.
- Außerwoger, J.; Ganglbauer, O.; Wallner, F.: Computerunterstützte
Fehlerdiagnose bei der Ultraschallprüfung von Schweißnahten.
Berg-Huttenmann. Monatsh. 130 (1985) 11, 417-421.
- Gebhardt, W.; Bonitz, F.; Woll, H.; Schmitz, V.:
Fehlercharakterisierung und Fehlerrekonstruktion mittels elektronischem
Sektorscan und Verbundscan. Saarbrucken, Fraunh. Ges. Inst. Zerstorungsfr.
Pruf., 1979, Bericht 790520-TW.
- Gebhardt, W.; Bonitz, F.; Woll, H.: Defect reconstruction and classification
by phased arrays. Mater. Eval. 40 (1982)1, 90-95.
- Wüstenberg, H.; Schulz, E.; Volker, J.; Sonnenberg, G.J.:
Turbinenteile uberwachen mit Ultraschall-Prufkopfen. MM Maschinenmarkt 90
(1984) 55/56, 1308-1311.
- Barbian, O.A.; Grohs, B.; Licht, R.:
Signalanhebung durch Entstorung von Laufzeitmeßwerten aus
Ultraschallprufungen von ferritischen und austenitischen Werkstoffen - ALOK 1.
Materialpruf. 23 (1981 ) 11, 379-383.
- Grohs, B.; Barbian, O.A.; Kappes, W.; Paul, H.:
Fehlerbeschreibung nach Art, Lage und Dimension mit Hilfe von
Laufzeitortskurven aus Ultraschallprufungen - ALOK 2.
Materialprüf. 23 (1981) 12, 427-432.
- Kappes, W.; Grohs, B.; Barbian, O.A.: Berechnung von Laufzeitortskurven
fur die Tandemprufung mit Ultraschall und Fehlerlage-Rekonstruktion aus Tandem-
Laufzeit-Ortskurven beim ALOK-Verfahren - ALOK 3. Materialprüf. 24 (1982) 5, 161-165.
- Crostack, H.A.; Roye, W.: Verbesserung der Ultraschallprüfung
von GuBteilen durch Anwendung
der Mehrfrequenzentechnik. GieBereiforsch. 36 (1984) 3,115-120.
- Crostack, H.A.; Roye, W.: Einsatz der Ultraschall-Mehrfrequenzentechnik
zur verbesserten Fehlerbeschreibung bei der Schweißnahtprüfung. Düsseldorf, Deut.
Verb. SchweiBtechn., 1986, DVS- Berichte 106, 37-41
- Vadder, D. de; Azou, P.; Saglio, R.; Birac, A.M.: Determination of orientation and
size of badly oriented quasi plane defects by means of focused probes.
Proceed. 9th WCNDT, Melbourne (1979), report 4G-1.
- Silk, M.G.: Defect sizing using ultrasonic diffraction.
Brit. J. Nondestr. Test. 21 (1979) 1,12-15.
- Silk, M.G.: A time approach to crack location and sizing in austenitic welds.
Brit. J. Nondestr. Test. 22 (1980) 2, 55-61.
- Schlengermann, U.: Zur Bestimmung von Rißtiefen durch Ultraschallverfahren - ein Überblick.
DGZfP-Jahrestagung 1987, Berichtsband 10, 70-81, auch Krautkrämer SD 264.
- Lam, F.K.; Tsang, W.M.: Flaw characterization based on diffraction of
ultrasonic waves. Ultrasonics 23 (1985) 1,14-20.
- Proegler, H.: Recent ultrasonic methods for the distinction between different types
of defects in welds. Proceed. 9th WCNDT, Melbourne (1979), report 3B-3.
- Charlesworth, J.P.; Temple, J.A.G.: Creeping waves in ultrasonic nondestructive testing.
Proceed. Ultrasonics International (1981),390-395.
- Budenkov, G.A.; Khakimova, L.I.: Measurement of the dimensions of spherical
and cylindrical defects. Sov. J. Nondestr. Test.17 (1981) 7, 540-545.
- Ogura, Y.; Ishii, Y.: A measurement of the height of internal defects using
the mode-converted surface wave at the defect. Proceed. 9th WCNDT,
Melbourne (1979), report 4G-5.
- Sachse, W.; Golan, S.: The scattering of elastic pulses and the nondestructive
evaluation of materials. ASME report ADM-29 (1978) 11-31.
- Gruber, G.J.: Defect identification and sizing by the ultrasonic satellite
pulse technique. J. Nondestr. Eval.1 (1980) 4, 263-276.
- Gebhardt, W.; Walte, F.: Ultraschallprufung auf senkrecht orientierte
Risse durch Ausnutzung der Wellenumwandlung (Tandem-ersatzprufung).
Materialpruf. 30 (1988) 3, 73-77.
- Gruber, G.J.; Hamlin, D.R.; Grothues, H.L.; Jackson, J.L.:
Imaging of fatigue cracks in cladded pressure vessels with the SLIC-50.
Nondestr. Test. Int.19 (1986) 3, 155-161.
works many years in the development and application department
for Krautkrämer GmbH D-Hürth.
This Article is a 'reprint' and was published in the Echo 33,34,35.
"the echo" is published free of charge at various intervals. Produced by:
Krautkrämer GmbH Co.
P. 0. Box 13 63 D-50354 Hürth
Contact: Krautkraemer in NDT online exhibition
© Copyright Rolf Diederichs 01. Dec 1996
/DB:Article /AU:Schlengermann_U /IN:Krautkramer /CN:DE /CT:UT /CT:sizing /ED:1996-12