See discussions, stats, and author profiles for this publication at: [Link]

net/publication/233395768

Nondestructive Testing Ultrasonic Immersion Probe Assessment and

Uncertainty Evaluation According to EN 12668-2:2010

Article  in  IEEE transactions on ultrasonics, ferroelectrics, and frequency control · October 2012

DOI: 10.1109/TUFFC.2012.2459 · Source: PubMed

CITATIONS READS

7 615

3 authors:

Cristiane E R Silva Andre Victor Alvarenga

Instituto Nacional de Tecnologia National Institute of Metrology, Quality and Technology (Inmetro)
20 PUBLICATIONS   64 CITATIONS    107 PUBLICATIONS   663 CITATIONS   

SEE PROFILE SEE PROFILE

Rodrigo P B Costa-Felix
National Institute of Metrology, Quality and Technology (Inmetro)
95 PUBLICATIONS   318 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Characterization of microcalcifications in mammograms View project

Interdisciplinary project for the study of bone demineralization View project

All content following this page was uploaded by Cristiane E R Silva on 06 June 2017.

The user has requested enhancement of the downloaded file.

2338 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control , vol. 59, no. 10, October 2012

Correspondence
Nondestructive Testing Ultrasonic Immersion maximum permissible errors (sometimes wrongly called
Probe Assessment and Uncertainty tolerances, according to [3]).
Evaluation According to EN 12668-2:2010

Cristiane E. R. Silva, André V. Alvarenga,

II. Beam Parameters From EN 12668-2:2010
and Rodrigo P. B. Costa-Felix
The beam parameters are determined by mapping the
Abstract—Ultrasonic beam parameters from nondestructive ultrasonic beam in water, using a reflector target or hy-
testing probes and respective measurement uncertainties were
drophone receiver. Within this paper, a hydrophone is
calculated according to the Guide to the Expression of Uncer-
tainty in Measurement (BIPM-JCGM-100:2008). The major selected because this ensures that signals have a better
conclusion of this work is that proper measurements of ultra- signal-to-noise ratio. The procedure comprises a beam ax-
sonic probes parameters are necessary because each probe has ial profile mapping to determine focal distance (FD) and
intrinsic construction particularities. Uncertainty evaluation focal length (FL), and also the position after the focal
was essential to properly assess the experimental results.
point where the axial profile is reduced by 3 dB, called
ZL2. Beam transverse profiles (on the x-axis and y-axis)
are mapped to establish focal widths at the focal point
I. Introduction (Wx1 and Wy1) and at ZL2 position (Wx2 and Wy2). Final-
ly, the divergence angles (Ωx and Ωy) are also calculated.
T o ensure the accuracy and repeatability of ultrasonic
tests, calibration of the nondestructive testing (NDT)
ultrasonic measurement system (measuring instrument,
All definitions according to EN12668-2 [1] can be found
elsewhere [2].
According to EN 12668-2:2010 [1], focal distance, focal
probe, standard block, etc) should be a common prac- length, and focal widths shall be within ±15% of the man-
tice. Ultrasonic probes play a key role in any ultrasonic ufacturer’s specifications, whereas the divergence angles
measurement system because they generate and receive shall not differ from values declared by the manufacture
the ultrasonic waves, regardless of whether the probe is by either ±10% or ±1°, whichever is the largest.
single- or multi-element.
The standard EN 12668-2:2010, Non Destructive Test-
ing—Characterization and Verification of Ultrasonic Ex- III. Materials and Methods
amination Equipment—Part 2: Probes, particularly Sec-
tion 7.7, Beam Parameters for Immersion Probes, defines The tests were performed using 10 different NDT ultra-
focal distance, focal length, focal widths, and beam di- sonic unfocused probes with nominal frequencies varying
vergence as the parameters that should be quantified to from 0.5 to 10 MHz. All probes had nominal diameters of
characterize the ultrasonic beam and ensure the system 12.7 mm except the 0.5-MHz frequency probe, which had
performance during an ultrasonic test [1]. a nominal diameter of 25.4 mm.
This work presents evaluation of beam parameters of The probes up to 5  MHz were excited by using a 30-cycle
ultrasonic immersion probes in accordance with EN12668- sine wave burst, whereas the 7.5- and 10-MHz probes were
2-2010 [1] and estimates respective measurement uncer- excited by using a 100-cycle sine wave burst, both gener-
tainty. The measurement system was developed at In- ated by an AFG 3252 function generator (Tektronix Inc.,
metro’s Laboratory of Ultrasound (Labus) [2]. Based on Beaverton, OR). The waterborne signals were acquired
the proposed measurement system, ultrasonic unfocused using needle hydrophones with 0.5-mm active elements
immersion probes in the range of 0.5 to 10 MHz were as- (Precision Acoustics Ltd., Dorchester, Dorset, UK), and
sessed. The obtained values for the various beam param- a TDS 3032B oscilloscope (Tektronix Inc.). A positioning
eters were statistically compared with the manufacturer’s system (Newport Corp., Irvine, CA) was used to move the
declared specifications, taking into account EN12668-2’s hydrophone in the water bath. All system components are
integrated by a virtual instrument (VI) developed in Lab-
VIEW (National Instruments Corp., Austin, TX), which
Manuscript received February 6, 2012; accepted July 2, 2012. The
authors acknowledge financial support from the Fundação Carlos Chagas provides a user-friendly interface. More details about this
Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) system can be found in [2] and [4].
through grant number E-26/190.037/2012 and the National Council for The measurement procedure starts by aligning the
Scientific and Technological Development (CNPq) through grant num-
bers 563166/2010-0, 385032/2010-2, 310414/2011-3, and 303769/2011-4. probe and hydrophone. Because the EN 12668-2:2010
The authors are with the Laboratory of Ultrasound, Directory of Sci- standard does not specify an alignment procedure, the
entific and Industrial Metrology (DIMCI), National Institute of Metrol- alignment was performed according IEC 62127-1 Ultra-
ogy, Quality and Technology (INMETRO), Duque de Caxias, RJ, Brazil
(e-mail: cesilva@[Link]). sonics—Hydrophones—Part 1: Measurement and Charac-
DOI [Link] terization of Medical Ultrasonic Fields up to 40 MHz [5],

0885–3010/$25.00 © 2012 IEEE

silva et al.: NDT ultrasonic immersion probe assessment and uncertainty evaluation 2339

which defines the beam-alignment axis as a straight line tenuation of the plane wave in the medium. The attenua-
joining two points of spatial-peak temporal-peak acoustic tion coefficient was estimated based on water temperature
pressure on two plane surfaces parallel to the face of the (T) by [8]
ultrasonic transducer. For this purpose, scans were per-
formed over the transverse profile (y-axis) at distances of α = (5.68524 × 10 1 − 3.02545T + 1.17416 × 10 −1T 2
A/(πλ) and A/(3πλ), where A is the nominal value of the − 2.95430 × 10 −3T 3 + 3.96985 × 10 −5T 4 (3)
geometrical area of the ultrasonic transducer and λ is the −7 5 −15 −2 −1
wavelength of the ultrasound corresponding to the nomi- − 2.11091 × 10 T ) × 10 [Hz ⋅m ].
nal value of the acoustic working frequency. This proce-
dure was repeated until the beam-alignment axis criterion The attenuation correction as shown in (2) was only
has been achieved. applied when comparing calculated values to those de-
Afterward, a scan along the beam axis using steps of clared within the manufacturer’s manual or specifications.
1.0 mm is carried out to determine the focal distance posi- However in practical field work, no attenuation correction
tion (FD). Then, the probe is mapped over two transverse would be applied because the measured values reproduce
axes to beam axis (X and Y), at the FD position, in a the practical work situation itself.
dimension sufficient to obtain signals lower than 10 dB of The probe effective diameter (deff) is a quantity de-
the amplitude value found at FD. The same two transverse rived from the ultrasound signal frequency, the propaga-
axes are mapped at the ZL2 position. In both procedures, tion medium, and the distance from the source to the
the steps used are 0.1 mm for probes up to 5 MHz, and target (hydrophone or reflecting target, for instance). This
0.05 mm for probes up to 10 MHz. parameter is closely related to the position of the last
All recorded mappings are uploaded using the VI and maximum and the last minimum of pressure along the
the values of focal distance (FD), focal length (FL), focal axis of symmetry perpendicular to, and outbound from,
width on FD (Wx1 and Wy1), focal width on ZL2 (Wx2 and the transducer surface. The deff,
Wy2), and divergence angles (Ωx and Ωy) are calculated
d eff = 2 ⋅ 2 ⋅ λ ⋅ z min + λ 2, (4)
and presented. The complete measurement procedure is
repeated 4 times, and at each repetition, the probe is re-
moved and replaced at the measurement system in a dif- was calculated based on the last axial minimum (zmin),
ferent position. (The probe is randomly rotated over its because this position provides a better spatial definition,
own beam axis.) and it is not affected by the attenuation [9].
The performance of the system was evaluated in terms For each set of results, an attenuation correction was
of FD, FL, Wx1, Wx2, Wy1, Wy2, and Ωx and Ωy values, applied to the measured FD values, leading to a corrected
as well as the respective measurement uncertainty from last maximum (FD_cor). Moreover, another last maximum
the studied NDT probes. The results were compared with was assessed from deff, denoted FD_eff, replacing dnominal in
the manufacturer’s declared specifications [6], [7] by cal- (5) by deff. Both FD_eff and FD_cor were reported to allow
culating the normalized error (1), taking into account the inter-comparison.
specified tolerance from EN 12668-2:
IV. Beam Parameters from
xA − xB Manufacture’s Specification
Errornormalized = , (1)
U x2A + U x2B
The technical specifications for probes used in this pa-
per are declared in technical notes published by their re-
where xA represents the value of one parameter measured,
spective manufacturers. Probes are considered to be work-
xB represents the value of the same parameter declared by
ing as a plane circular piston, and NDT probe users take
the manufacturer, U x2A is the measurement uncertainty,
into account the technical notes to estimate beam param-
and U x2B represents the specified tolerance from EN 12668- eters and specify which kind of transducer will be appro-
2. priate for each task. Among the parameters presented in
For frequencies higher than 5 MHz, the effects of at- technical notes and studied here, FD and FL values for an
tenuation on ultrasonic propagation become significant unfocused probe are estimated by (5) and (6), respectively
and a distance-dependent correction should be taken into [6], [7]:
account. Hence, when comparing the measured values
with the declared values, the ultrasonic pressure along the 2
d nominal  2
beam axis probe was corrected as [8] FD =  1 −  λ   , (5)
4λ   d 
nominal  
P(z ) 4
Pcorrected = 2 , (2) FL = FD, (6)
e −αzf 3

where P(z) is the pressure values obtained during the scan where dnominal is the nominal probe diameter and λ is the
along the beam axis probe, z is the distance along the wavelength, obtained by the relation between the opera-
Z-axis, f is the probe operation frequency, and α is the at- tion frequency probe ( f ) and the ultrasonic velocity in the
2340 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control , vol. 59, no. 10, October 2012

TABLE I. Parameter Values and Respective Uncertainty Values for NDT Probe 1
(0.5 MHz, Diameter = 25.4 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 53.60 58.3 3.4 5.8 0.533
FL [mm] 71.50 90.5 5.9 6.6 1.549
Wx1 [mm] — 8.13 0.66 8.1 —
Wy1 [mm] — 8.28 0.49 5.9 —
Wx2 [mm] — 15.08 0.63 4.2 —
Wy2 [mm] — 15.4 1.3 8.5 —
Ωx [°] 3.44 2.92 0.31 10.5 0.457
Ωy [°] 3.44 2.97 0.21 7.1 0.422

TABLE II. Parameter Values and Respective Uncertainty Values for NDT Probe 2
(1 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 26.80 27.75 0.80 2.9 0.232
FL [mm] 35.80 40.00 0.82 2.1 0.773
Wx1 [mm] — 3.68 0.10 2.7 —
Wy1 [mm] — 3.70 0.09 2.4 —
Wx2 [mm] — 6.80 0.20 2.9 —
Wy2 [mm] — 6.85 0.25 3.6 —
Ωx [°] 3.44 3.06 0.30 9.8 0.338
Ωy [°] 3.44 3.08 0.27 8.8 0.319

TABLE III. Parameter Values and Respective Uncertainty Values for NDT Probe 3
(2.25 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 61.00 60.50 0.87 1.4 0.054
FL [mm] 81.40 83.0 4.3 5.1 0.124
Wx1 [mm] — 3.40 0.13 3.8 —
Wy1 [mm] — 3.45 0.15 4.3 —
Wx2 [mm] — 6.55 0.35 5.3 —
Wy2 [mm] — 6.60 0.17 2.6 —
Ωx [°] 1.53 1.443 0.063 4.4 0.174
Ωy [°] 1.53 1.443 0.063 4.4 0.174

TABLE IV. Parameter Values and Respective Uncertainty Values for NDT Probe 4
(2.25 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 60.90 63.0 2.9 4.6 0.208
FL [mm] 81.20 93.75 0.97 1.0 1.008
Wx1 [mm] — 3.88 0.29 7.5 —
Wy1 [mm] — 3.93 0.37 9.4 —
Wx2 [mm] — 7.00 0.19 2.8 —
Wy2 [mm] — 7.08 0.18 2.5 —
Ωx [°] 1.53 1.283 0.056 4.4 0.507
Ωy [°] 1.53 1.293 0.065 5.0 0.487
silva et al.: NDT ultrasonic immersion probe assessment and uncertainty evaluation 2341

TABLE V. Parameter Values and Respective Uncertainty Values for NDT Probe 5
(3.5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 95.20 97.8 2.8 2.8 0.175
FL [mm] 127.00 135.3 2.1 1.5 0.430
Wx1 [mm] — 3.700 0.082 2.2 —
Wy1 [mm] — 3.68 0.18 4.9 —
Wx2 [mm] — 6.80 0.13 1.9 —
Wy2 [mm] — 6.90 0.13 1.9 —
Ωx [°] 0.98 0.886 0.048 5.4 0.302
Ωy [°] 0.98 0.921 0.053 5.8 0.189

medium (c). It is worth emphasizing that the ultrasonic mated as the standard deviation from 4 repeated measure-
velocity was estimated based on the temperature of the ments, divided by 4. Finally, the Ωx and Ωy uncertainties
water [10], because tests were performed in that medium. take into account uncertainties obtained from each of the
Similarly, the values of Ωx = Ωy can be determined by [6], parameters used in beam divergence calculation. Finally,
[7] the expanded uncertainty was calculated considering a
coverage factor k = 2 (95.45% confidence level).
 c 
Ω x = Ω y = arcsin  0.514 * . (7)
 f ⋅ d nominal 
VI. Results
Herein, the parameter values calculated from (5), (6),
and (7) presented in manufactures’ technical notes [6], [7], The beam parameters (FD, FL, Wx1, Wx2, Ωx, and Ωy)
and the tolerance ranges defined in EN 12668-2 will be were measured on ten 12.7-mm-diameter probes (one
considered as the declared manufacture’s specifications to 1-MHz, two 2.25-MHz, one 3.5-MHz, two 5-MHz, two
allow inter-comparison with measured values. 7.5-MHz, and two 10-MHz probes), and one 25.4-mm-
diameter probe (0.5 MHz). The results are presented in
Tables I–XI.
V. Determination of Standard For frequencies higher than 5 MHz, the beam axis pro-
Uncertainty of Type A and Type B files determined after attenuation correction can be seen
in Fig. 1, whereas the new parameter values, also cal-
All parameters except beam divergence are calculat- culated after the attenuation correction, can be seen in
ed directly from the raster scans of axial and transversal Tables XII–XV. Finally, the values FD_eff, deff, and zmin
beam profiles. Hence, the raster scan resolution and the are presented in Table XVI.
step size (s) are the principal sources of uncertainty for
FD, FL, Wx1, Wx2, Wy1, and Wy2. All uncertainties were
assessed according to [11]. VII. Discussion
The positioning system resolution is assumed to pres-
ent a rectangular distribution; hence, the Type B uncer- Considering the NDT ultrasonic immersion probes in
tainty is estimated by dividing the resolution by 2 3. Step the range of 0.5 to 5 MHz studied here (Tables I–VII), the
size is also assumed to have a rectangular distribution, measured FD values can be accepted using the EN 12668-
and thus its Type B uncertainty is defined as the used step 2 criterion (±15% of the manufacturer’s specifications),
size divided by 2 3. The Type A uncertainties were esti- because all normalized errors were lower than 1. Similarly,

TABLE VI. Parameter Values and Respective Uncertainty Values for NDT Probe 6
(5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 136.00 127.8 3.0 2.3 0.400
FL [mm] 181.00 154.0 1.2 0.8 0.993
Wx1 [mm] — 3.03 0.14 4.6 —
Wy1 [mm] — 3.05 0.11 3.6 —
Wx2 [mm] — 5.425 0.097 1.8 —
Wy2 [mm] — 5.38 0.14 2.6 —
Ωx [°] 0.69 0.628 0.043 6.8 0.266
Ωy [°] 0.69 0.608 0.044 7.2 0.356
2342 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control , vol. 59, no. 10, October 2012

TABLE VII. Parameter Values and Respective Uncertainty Values for NDT Probe 7
(5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 136.00 142.3 2.1 1.5 0.305
FL [mm] 181.00 173.0 1.9 1.1 0.294
Wx1 [mm] — 3.55 0.11 3.1 —
Wy1 [mm] — 3.50 0.13 3.7 —
Wx2 [mm] — 6.40 0.13 2.0 —
Wy2 [mm] — 6.375 0.097 1.5 —
Ωx [°] 0.69 0.653 0.028 4.3 0.092
Ωy [°] 0.69 0.650 0.030 4.6 0.065

TABLE VIII. Parameter Values and Respective Uncertainty Values for NDT Probe 8
(7.5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 204.00 163.5 6.7 4.1 1.290
FL [mm] 272.00 163.3 4.9 3.0 2.644
Wx1 [mm] — 2.68 0.10 3.7 —
Wy1 [mm] — 2.63 0.13 5.0 —
Wx2 [mm] — 4.350 0.062 1.4 —
Wy2 [mm] — 4.30 0.20 4.7 —
Ωx [°] 0.459 0.430 0.025 5.8 0.193
Ωy [°] 0.459 0.430 0.016 3.6 0.199

TABLE IX. Parameter Values and Respective Uncertainty Values for NDT Probe 9
(7.5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 204.00 191.5 8.3 4.3 0.391
FL [mm] 272.00 196.8 4.9 2.5 1.828
Wx1 [mm] — 3.38 0.22 6.5 —
Wy1 [mm] — 3.475 0.051 1.5 —
Wx2 [mm] — 5.41 0.19 3.5 —
Wy2 [mm] — 5.45 0.15 2.8 —
Ωx [°] 0.459 0.418 0.013 3.0 0.276
Ωy [°] 0.459 0.406 0.038 9.4 0.350

TABLE X. Parameter Values and Respective Uncertainty Values for NDT Probe 10
(10 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 272.00 196.8 9.7 4.9 1.792
FL [mm] 362.00 164.8 2.3 1.4 3.633
Wx1 [mm] — 2.49 0.12 4.8 —
Wy1 [mm] — 2.46 0.11 4.5 —
Wx2 [mm] — 3.700 0.080 2.2 —
Wy2 [mm] — 3.700 0.080 2.2 —
Ωx [°] 0.344 0.318 0.018 5.7 0.233
Ωy [°] 0.344 0.325 0.027 8.3 0.172
silva et al.: NDT ultrasonic immersion probe assessment and uncertainty evaluation 2343

TABLE XI. Parameter Values and Respective Uncertainty Values for NDT Probe 11
(10 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 272.00 231.3 5.4 2.3 0.997
FL [mm] 362.00 243.3 4.3 1.8 2.194
Wx1 [mm] — 2.83 0.22 7.8 —
Wy1 [mm] — 2.88 0.19 6.7 —
Wx2 [mm] — 4.13 0.14 3.4 —
Wy2 [mm] — 4.14 0.14 3.4 —
Ωx [°] 0.344 0.330 0.026 7.9 0.121
Ωy [°] 0.344 0.320 0.020 6.2 0.210

TABLE XII. Parameter Values and Respective Uncertainty Values After Attenuation

Correction for NDT Probe 8 (7.5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 203.88 170.8 8.0 4.7 1.048
FL [mm] 271.84 213.5 9.6 4.5 1.393
Ωx [°] 0.459 0.423 0.020 4.7 0.193
Ωy [°] 0.459 0.429 0.012 2.8 0.199

TABLE XIII. Parameter Values and Respective Uncertainty Values After Attenuation

Correction for NDT Probe 9 (7.5 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 203.87 212 10 4.7 0.237
FL [mm] 271.83 271.3 6.5 2.4 0.014
Ωx [°] 0.459 0.4211 0.0073 1.7 0.276
Ωy [°] 0.459 0.395 0.026 6.6 0.350

TABLE XIV. Parameter Values and Respective Uncertainty Values After Attenuation

Correction for NDT Probe 10 (10 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit value Result uncertainty [%] error
FD [mm] 271.83 224.8 6.3 2.8 1.145
FL [mm] 362.44 267.8 8.8 3.3 1.713
Ωx [°] 0.344 0.318 0.013 4.1 0.233
Ωy [°] 0.344 0.325 0.026 8.0 0.172

TABLE XV. Parameter Values and Respective Uncertainty Values After Attenuation

Correction for NDT Probe 11 (10 MHz, Diameter = 12.7 mm).
Expanded
Declared Expanded uncertainty Normalized
Parameter Unit values Results uncertainty [%] error
FD [mm] 272.35 257 14 5.5 0.342
FL [mm] 363.14 308.3 3.0 1.0 0.988
Ωx [°] 0.343 0.330 0.025 7.6 0.121
Ωy [°] 0.343 0.320 0.016 5.0 0.210
2344 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control , vol. 59, no. 10, October 2012

Fig. 1. Beam axis profiles from probes (a) 8, (b) 9, (c) 10, and (d) 11 after attenuation correction.

the measured beam divergence parameters (Ωx and Ωy) and 4 (2.25 MHz, diameter = 12.7 mm) were significantly
can also be considered acceptable (either ±10% or ±1° different from the declared values. Considering that NDT
whichever is the largest). However, the measured FL val- probe 3 has the same frequency and diameter of NDT
ues from NDT probes 1 (0.5 MHz, diameter = 25.4 mm) probe 4 and presented adequate values of FD, it was de-

Fig. 2. Ultrasonic pressure variation along the axis of ultrasonic beam in water for NDT probes (a) 3 and (b) 4, both of frequency 2.25 MHz and
diameter = 12.7 mm; (c) and (d) are images of planar scanning at 3 mm from the faces of the probes 3 and 4, respectively.
silva et al.: NDT ultrasonic immersion probe assessment and uncertainty evaluation 2345

TABLE XVI. FD Values Calculated After Attenuation Corrections (FD_cor), FD_eff Values
Estimated Based on the Probe Effective Diameter (deff), Values of the Last Axial
Minimum (zmin), and Probe Nominal Diameters (dnominal).
Frequency zmin dnominal deff FD_cor FD_eff
Probe [MHz] [mm] [mm] [mm] [mm] [mm]
8 7.5 87 12.7 11.7 170.8 174.2
9 7.5 104.5 12.7 12.9 211.5 209.1
10 10 110.8 12.7 11.5 224.8 221.7
11 10 128.8 12.7 12.4 257.3 257.7

cided to compare the ultrasonic pressure along the beam 11: Tables IX and XI, respectively). As previously men-
axis from both probes (Fig. 2). One can observe that the tioned, the ultrasonic pressure along the beam axis was
ultrasonic pressure from NDT probe 4 tends to present a corrected for frequencies higher than 5 MHz, and after
slower slope after the last maximum than the one from correction, the FL measured values from NDT probes 9
NDT probe 3, driving to higher values of FL. and 11 achieved acceptable results (Table XIII and Table
Considering the studied NDT ultrasonic immersion XV, respectively). However, even after attenuation cor-
probes in the range of 7.5 to 10 MHz studied here (Tables rection, one can observe that the values of FL and FD
VIII–XI), one can observe that all of the measured beam obtained from NDT probes 8 and 10 were not in accor-
divergence parameters (Ωx and Ωy) can be considered ac- dance with manufactures declared values. Analyzing the
ceptable using the EN 12668-2:2010 acceptance criterion. results presented in Table XVI, one can observe that NDT
However, none of the measured FL values were acceptable, probes 8 and 10 have effective diameters 7.9% and 9.4%
and only two FD values were adequate (NDT probes 9 and smaller than the nominal diameter, whereas the difference

Fig. 3. Mappings over the (a) x-axis and (b) y-axis parallel to the transducer face at position ZP, and the (c) x-axis and (d) y-axis at position ZL2 for
the probe 10. Note that profiles are different among repetitions, suggesting that the transducer does not have the behavior of a theoretical circular
piston.
2346 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control , vol. 59, no. 10, October 2012

between nominal and effective diameters for NDT probes Once it has been shown that there are differences be-
9 and 11 are 1.6% and 2.4%, respectively. Furthermore, tween what is presented in a technical note delivered by
as previously mentioned, at each repetition the probe is the manufacturer and the real transducer output, and as
randomly turned around over its own beam axis. Under far as it is offered as complementary (or fundamental) in-
the assumption that the probe is acting as a plane pis- formation, one should use that information to choose (or
ton, it was expected that the mapped transverse profile at reject) a determined transducer for a specific application.
each repetition should be the same. However, this was not Further, to properly identify an experimental param-
observed for some NDT probes, as one can see from the eter, uncertainties should be addressed. The present work
scans carried out at the ZP and ZL2 beam profile distances could also be used as a general guide to assess uncertain-
in the four performed tests for probe 10 (Fig. 3). ties for ultrasonic probes measurement, particularly for
The observed differences are probably related to the those to be used in nondestructive testing.
construction characteristics of each probe, for example,
the process by which the piezoelectric ceramic is affixed
to the backing layer, as well as the position and manner References
of electrode attachment, the uniformity and homogene-
ity of front face matching layers, and the quality of bond [1] Non Destructive Testing—Characterization and Verification of Ul-
between ceramic element and backing. These construc- trasonic Examination Equipment—Part 2: Probes, EN 12668-2:2010,
2010.
tion characteristics might produce different experimental [2] C. E. R. Silva, A. V. Alvarenga, and R. P. B. Costa-Félix, “Ultra-
and theoretical behaviors. These results are important, sonic immersion probes characterization for use in nondestructive
because they suggest that industrial construction is not a testing according to EN 12668-2:2001,” J. Phys.: Conf. Ser., vol.
279, pp. 1–6, 2010.
guarantee that probes of the same model will work simi- [3] International Vocabulary of Metrology—Basic and General Concepts
larly. Obviously, dependent on the parameter values, such and Associated Terms (VIM), BIPM JCGM 200:2008, 2008.
differences do not prevent or make the probe utilization [4] A. V. Alvarenga and R. P. B. Costa-Félix, “Uncertainty assess-
ment of effective radiating area and beam non-uniformity ratio of
unfeasible for any given application. However, these re- ultrasound transducers determined according to IEC 61689:2007,”
sults point out the need for a specific test report for each Metrologia, vol. 46, no. 3, pp. 367–374, 2009.
manufactured probe unit. [5] Ultrasonics—Hydrophones—Part 1: Measurement and Characteriza-
tion of Medical Ultrasonic Fields up to 40 MHz, IEC 62127-1:2007,
2007.
VIII. Conclusion [6] Olympus NDT Inc. (2011, Jul.) Ultrasonic Transducers Techni-
cal Notes. [Online]. Available: [Link]
.downloads/download/?file=285213010&fl=en_US
Nondestructive testing comprises a set of measure- [7] GE Sensing & Inspection Technologies. (2011, Jul.) Ultrasonic
ments for which trustworthiness is deeply dependent on Transducers for Flaw Detection and Sizing. [Online]. Available:
[Link]
the behavior of the instruments. The probes are one of the 20117EN_ultrasonic-[Link]
system components that are more susceptible to variation [8] Ultrasonics—Hydrophones—Part 2: Calibration for Ultrasonic Fields
as function of construction constraints. Theoretical pre- up to 40 MHz, IEC 62127–2:2007, 2007.
[9] R. T. Hekkenberg, “Hydrophone free-field open-circuit sensitivity
diction of ultrasonic probe’s behavior varies significantly contribution of TNO PG to the BIPM/CIPM Key Comparison
from one unit to other, due mostly to their assembly. Be- US2,” TNO research report, PG/TG/00.029, 1980.
cause of that, final users should have the most important [10] National Physical Laboratory, “Underwater Acoustics, Technical
Guides—Speed of sound in pure water,” NPL, Teddington, Middle-
parameters of ultrasonic probes determined experimen- sex, UK, 2000.
tally instead of just using theoretical formulations. The [11] Evaluation of Measurement Data—Guide to the Expression of Uncer-
present work reinforces that concept. tainty in Measurement (GUM), BIPM JCGM 100:2008, 2008.

View publication stats