ECNDT 2006 - Poster 172

Application of ISO 5725-1-5725-6 for the

Validation of Ultrasonic Testing
Vera KONSHINA, Gregory DYMKIN, NIIM/PGUPS, St. Petersburg, Russia
Christina MUELLER, Federal Institute for Materials Research and Testing (BAM), Berlin,
Germany

Abstract. On the basis of main provisions of ISO 5725-1-5725-6 standards,

reproducibility and repeatability in determination of measured characteristics of
artificial reflectors and real defects have been evaluated, and suggestions are given
for including evaluation of repeatability and reproducibility of measured
characteristics in the ultrasonic testing technique (and equipment) validation
procedure.

Introduction and statement of problem

During recent decades, the problem of assessment ("validation", "verification",

"qualification" …) of non-destructive testing (NDT) techniques attracts close attention of a
wide circle of specialists – both developers and "users" of testing techniques. Most
researchers estimate quality of techniques by their "trustworthiness" which is characterized
by a number of indicators analyzed in detail in [1, 2].
The most complete and accurate assessment of trustworthiness for qualification of
an ultrasonic testing technique can be obtained from results of testing a sufficiently large
quantity of specimens followed by determining the real defect situation by means of
destruction of specimens and analysis of opened cross-sections of the specimens under
testing. Naturally, such qualification procedure is practically unacceptable. There exists a
possibility of using specimens with artificial reflectors covering the entire range of
reflective properties of defects. However, in this case, some uncertainty of the actual defect
situation exists since it is characterized by values of measured characteristics of reflectors,
determination of which demonstrates significant data spread.
International standards of ISO 5725 series [3, 4] regulate methods for assessment of
accuracy in measurements of various physical values in compliance with a standardized
procedure (technique) and introduce concepts of reproducibility and repeatability of
measurement results.

Main provisions of ISO 5725 standards in terms of NDT

The term "accuracy" when it is related to a series of measurement or test results includes a
combination of random components and a general bias [5].
In the process of ultrasonic testing, measured characteristics of defects are
determined [6]. A decision on acceptability or unacceptability of defects detected during
testing is made by comparing the values of measured characteristics of defects to rejection
criteria, values of which are preset also by normative documents for testing. Therefore, the

1
correct assessment of fitness of the tested object and classification is influenced by the
accuracy of defects characteristics measurement.
ISO 5725-1 standard introduces the concept of:
• "correctness" understood as a degree of closeness of a measurement result to
the true (actual) value of the quantity being measured, which can be adopted
[3] as a mean value of a preset collection of measurement results;
• "precision" which is considered as a degree of closeness between
independent measurement results obtained in particular established
conditions and depends on random factors only. The measure of precision is
usually calculated as standard (root-mean-square) deviation of results of
measurements performed in definite conditions.

Also introduced are concepts of "repeatability" and "reproducibility", which are two
extreme cases of precision, where the former concept characterizes minimum and the latter
maximum variability of results.
In practical interpretation of measurement results, this variability must obligatorily
be taken into account. For instance, for assessment of ultrasonic testing results, the obtained
value of the measured characteristic of a detected defect (amplitude of the echo signal from
the defect, equivalent area, conventional extent, etc.) is compared to the threshold value
established by the relevant normative document [6]. However, the actual difference
between the obtained result of measurements and the preset value, of course, cannot be
established if they both lie in the region of inevitable random errors of the measurement
procedure.
In consideration of difficulties in assessment of trustworthiness of non-destructive
testing techniques, and taking into account that application of the reproducibility and
repeatability concepts is possible even if information about the actual value of the measured
quantity is unavailable, an attempt is made in this report to use them for comparative
assessment of ultrasonic testing techniques.

Evaluation of precision in determination of measured characteristics of artificial

reflectors and real defects

The most important measured characteristics of defects, in accordance with various

normative documents, include: defect coordinates, maximum echo signal amplitude (or
detectability factor), conventional extent measured by various methods. The detectability
factor Kd is the difference, in dB, between amplitude of the echo signal from the defect and
amplitude of the echo signal from a certain artificial reflector.
In order to assess precision of determining the depth of artificial reflectors location,
results of measurement of lateral cylindrical holes were used. The measurements were
carried out by operators of the same level of skills. Flaw detectors of the same type with
incidence angles of 50о and 65о were used. The results of precision assessment in
determining the depth of artificial reflectors location are presented in Figure 1.
It is known that the error in determination of defect location depth increases with
increasing the angle of incidence [6, 7]. Accordingly, worse reproducibility of reflectors
location depth measurement shown in Figure 1 can be attributed to an increased direction-
finding error when a transducer with a large angle of incidence is used.
In order to assess precision of determining the conventional extent and detectability
factors of artificial reflectors, we used results of tests on specimens that had models of

2
defects on their surfaces made in the form of vertical cylindrical holes from 2 to 4 mm in
diameter. The weld itself was simulated by a bar that, at the same time, hid existing
reflectors from the operator. Testing was carried out by two groups of operators. The first
group included operators without practical experience while operators of the second group
had working experience longer than 10 years. In the process of testing, the operators
performed scanning and determined measured characteristics of reflectors.
Each point in Figure 2 corresponds to results of conventional extent measurement of
a definite diameter hole averaged for a group of five operators. First of all, Figure 2
confirms that the average value of an undirected reflector conventional extent measured by
a relative method does not depend on the reflector diameter. Reproducibility of
conventional extent measurement results is, naturally, higher for experienced operators than
for those without work experience.
Accordingly, depending on the minimum conventional size of the defect preset by
the testing technique, and taking the precision evaluation data into account, a conclusion
can be made that it is possible to fulfill the technique requirements during testing, in the
given case, by operators having different skills. It follows from Figure 2 that, if minimum
conventional extent of an unacceptable defect is 15 mm, then the requirements of such
technique will be fulfilled even by low-skilled operators, and if it is 10 mm, then such
technique can be realized only by operators having experience in the work.
For studying reproducibility and repeatability of determining measured
characteristics of actual defects, three skilled operators measured detectability factors and
conventional extents of real defects of lack-of-fusion type (5 mm and 10 mm long) in butt-
welded joints with a direct beam (m = 0) and a singly reflected beam (m = 1) using
transducers with incidence angles of 65о and 50о, respectively.
Thus, determination of measured characteristics of defects can be quantitatively
characterized by reproducibility of results, which depends primarily on the operators' skills
and, in some cases, on testing parameters. It should be noted that the purpose of
experiments in reproducibility and repeatability determination is to find the limits within
which results may vary in practice, which is rather important information for establishing
rejection criteria by which the operator makes a conclusion about acceptability or
unacceptability of a detected defect.

3
 8

4
SR

2 angle of incidence 50 deg

angle of incidence 65 deg

0
0 20 40 60 80
Reflector depth, mm

Figure 1: Reproducibility of reflector depth measurements (11 operators 10 measurements each)

6 with work experience

5
without work experience
4
SR

0
0 2 4 6 8 10

Conventional extent, mm

Figure 2: Reproducibility of conventional extent measurements for vertical cylindrical holes (10 operators,
from 5 to 8 measurements)

Assessment of reproducibility as a tool for qualification of ultrasonic testing

techniques (equipment)

Thus, it has been shown that quantitative assessment of an ultrasonic testing technique
during its experimental study can be performed by determining reproducibility and/or

4
repeatability of determination of measured characteristics of defects to be detected in
compliance with requirements of NDT technique.
For experimental determining reproducibility and/or repeatability of determination
of measured characteristics of defects in compliance with requirements of NDT technique:
• specimens are prepared having real or artificial defects whose size
correspond to minimum characteristic sizes of defects to be detected;
• characteristics of defects in prepared specimens are measured;
reproducibility and repeatability of determination of measured characteristics
of real defects or reflector models are evaluated

The same procedure can be used for qualification of ultrasonic testing equipment.
Experimental check of the possibility to evaluate reproducibility of measurement
results for qualification of ultrasonic testing equipment has been carried out for ultrasonic
thickness gauges.
Determination of thickness of metal structures was performed by four operators of
the same skills using three types of thickness gauges. All measurements were performed 10
times on specimens of different thickness.
Figure 3 shows the difference in reproducibility of thickness measurements by
various instruments, which allows the justification of the selection of a particular thickness
gauge for inspection of elements having thickness in a definite range.

1,4

1,2

1
УД2-12
0,8
А1207С
SR

0,6 УТ-93Т
0,4

0,2

0
0 5 10 15 20 25 30
thickness, mm

Figure 3: Reproducibility of thickness measurements of metal structure components by thickness gauges of

various types

Conclusions

1. Determination of measured characteristics of defects can be characterized quantitatively

by reproducibility of results which is in compliance with the procedures described in
standards of ISO 5725 series.
2. The possibility of quantitative assessment of a technique (equipment) has been
demonstrated. The quantitative assessment of an ultrasonic testing technique during its
experimental study or during application can be carried out by determining the
reproducibility or repeatability of the determination of measured characteristics of

5
 defects to be detected with a particular NDT procedure or particular ultrasonic testing
equipment with affordable effort.

References

[1] Nockeman C.; Tillack G.-R.; Wessel H., Hobbs C., Konchina V.; Perfomance demonstration in NDT by
statistical methods: ROC and POD for ultrasonic and radiographic testing // 6th European Conference on
Non Destructive Testing, 24.-28. October 1994. – Nice, 1994. – Т.1. РР. 37-44
[2] Dymkin G.Y., Konshina V.N., Nokckeman C., Tillack G.-R.; On possibility of the use of operative
detection characteristics for trustworthiness assessment of non-destructive testing techniques. Theses of
reports of 15th Petersburg conference "Ultrasonic testing of metal structures"; St.Petersburg, 30 May – 01
June of 1995 – St.Petersburg, 1995.
[3] ISO 5725-1 Accuracy (correctness and precision) of measurement methods and results. Part 1.
Fundamentals and definitions.
[4] ISO 5725-2 Accuracy (correctness and precision) of measurement methods and results. Part 2. Basic
method for determining repeatability and reproducibility of standard measurement method.
[5] Dymkin G.Y., Konshina V.N., Nokckeman C., Tillack G.-R.; The use of thrust worthiness indicators for
validation of non-destructive testing techniques // Defectoscopia. – 2000 No. 3
[6] ISO 3534-1:1993 Statistics – vocabulary and designations – Part 1: Statistic methods. Terms and
definitions.
[7] Gurvich A.K., Ermolov I.N. Ultrasonic testing of welded joints. – Kiev: Technika, 1972, 460 pp.