CCM Vickers Key Comparison Final Report: Physikalisch-Technische Bundesanstalt Braunschweig

Physikalisch-Technische Bundesanstalt Braunschweig
CCM Vickers key comparison
Final Report
Braunschweig, March 2005 / K. Herrmann
1
Content
1 Introduction 3
2 Organisation 3
2.1 Participants 3
2.2 Time schedule 5
3 Standards 6
3.1 Description 6
3.2 Handling 7
4 Measurand 7
5 Methods of measurement 7
6 Stability of the standards 7
7 Measurement results 8
8 Analysis 12
8.1 Reference values on the basis of the uncertainty determined by the

participants 12
8.2 Reference values on the basis of a unified estimation of the hardness
measurement uncertainty 19
8.3 Influence of the inhomogeneity of hardness of the hardness reference
blocks 26
8.4 Influence of the length measuring deviation 27
8.5 Influence of the numerical aperture of the objective 31
9 Uncertainty budgets 32
9.1 Results of calibrations 22
9.2 Calculation of measurement uncertainty 35
9.3 Degree of equivalence 35
10 Discussions, conclusions and remarks 42
11 References 43
Abstract 44
Appendix:
A1 Description of the instruments by the participants
A2 Reference values
A2 Uncertainty budgets delivered by the participants
A3 Uncertainty budgets of the participants based on the unified procedure
A4 Calculation of the degree of equivalence
2
1 Introduction
The metrological equivalence of national measurement standards is determined by a

set of key comparisons chosen and organised by the Consultative Committees of the
CIPM or by the Regional Metrology Organisations (ROM's) in collaboration with the
Consultative Committees.
At its meeting in May 1999, the Working Group on Hardness (WGH) of the
Consultative Committee for Mechanical Measurements, CCM, identified several key
comparisons in the field of hardness measurements and decided upon the general
content.
In particular, the WGH has decided that a key comparison on the main Vickers
hardness scales should be carried out with the Physikalisch-Technische
Bundesanstalt (PTB) as the pilot laboratory.
As the rules for key comparisons were followed it may be declared as such and thus
be included in the MRA App. B data base.
2 Organisation
Following the rules set up by the BIPM1) a small group from the provisional list of
participating laboratories has drafted the technical protocol. The group was
composed of the pilot laboratory (Konrad Herrmann from PTB, Germany) and
Alessandro Germak from IMGC, Italy.
The draft of the technical protocol was agreed upon between the participants of the
comparison in March 2001. The comparison started in March 2001 and ended in
March 2003.
The participants were invited according to a questionnaire of the ad hoc Working
Group on Hardness about hardness laboratories in National Metrology Institutes from
June 1998.
2.1 Participants
The list of participants is given in Table 1.
___________________________
1)
T. J. Quinn, Guidelines for key comparisons carried out by Consultative
Committees, BIPM, Paris
3
Table 1: List of participants
Participant Institute Adress

Alessandro Germak Istituto di Metrologia "G. Colonnetti" Tel: +39 011 3977 367
Strada delle cacce, 73 Fax: +39 011 3977 426
10135 Torino e-mail: A.Germak@imgc.cnr.it
Italy
Renato Reis Machado National Institute of Metrology, Tel:+55 21 679 1311
Standardization and Industrial Fax:+55 21 679 1420
Quality e-mail: rreism@terra.com.br
Force & Hardness Laboratory
Av. N.S. das Gracas, 50 - Xerem
25250-020, Duque de Caxias, RJ,
Brasil
Samuel R. Low National Institute of Standards and Tel:001 301 975 5709
Technology Fax:001 301 975 4553
Materials Performance Group e-mail: samuel.low@nist.gov
Metallurgy Division
Materials Science and Engineering
Laboratory
100 Bureau Drive, Mail Stop 8553
Gaithersburg, Maryland 20899
U.S.A.
Hajime Ishida National Institute of Advanced Tel:+81 298 61 4048
Industrial Science and Technology Fax:+81 298 61 4047
Acoustics and vibration e-mail: h-ishida@aist.go.jp
measurement division
AIST Tsukuba Central 3
1-1-1, Umezono, Tsukuba
305-8563, Japan
He Li National Institute of Metrology Tel:+86 1 6421 8628
Mechanics Division Fax:+86 1 6421 8703
Hardness Laboratory e-mail: Liqznim@hotmail.com
18, Bei San Huan Dong Lu
100013 Beijing
P.R. China
Gun-Woong Bahng Korea Research Institute of Tel:+82 42 868 5320
Standards and Science Fax:+82 42 868 5032
Microstructure Analysis Group e-mail: gwbahng@kriss.re.kr
Materials Evaluation Center
P.O.Box 102
Yusong, Taejon 305-600
Republic of Korea
Jiri Borovsky Cesky Metrologicky Institut Tel:+420 2 5732 5579
Laboratore primarni metrologie Fax:+420 2 5732 8077
V botanice 4 e-mail: jborovsky@cmi.cz
15072 Praha
Czech Republic
Anna Osinska- Glowny Urzad Miar Tel:+22 6 200 241
Karczmarek Zaklad Masy i Sily Fax:+22 6 208 378
ul. Elektoralna 2 e-mail: mass.force@gum.gov.pl
00-950 Warszawa
Poland
Laurence Brice National Physical Laboratory Tel:+44 020 8943 6633
Force Standards Section Fax:+44 020 8943 6184
Centre for Mechanical and e-mail: laurence.brice@npl.co.uk
Acoustical Metrology
Queens Road
Teddington, Middlesex
TW11 0LW
United Kingdom
4
Coordinator:
Konrad Herrmann Physikalisch-Technische Tel:+49 531 592 5140
Bundesanstalt Fax:+49 531 592 5105
Project Hardness Measurement e-mail: konrad.herrmann@ptb.de
Bundesallee 100
38116 Braunschweig
Germany
2.2 Time schedule and measurements carried out by the participants

The comparison started in February 2001 with the initial measurements at PTB. Due
to problems with the availability and necessary repair of the hardness standard
machines at several participants the foreseen period of measurement for each
partner of one month could not be kept. Moreover, during the transport of the
hardness reference blocks as standards for the measurements the accompanying
customs documents were lost, so that the pilot laboratory had to care about this
customs problem. Therefore, considerable delay in the measurements occurred so
that the originally scheduled measurement period for the whole comparison of one
year was not realistic.
The following table shows the scheduled measuring time, the date of reception of the
standards and the date when the results were received by the pilot laboratory.
Table 2: Time schedule
Laboratory Original schedule Confirmation of Results received

reception
PTB 02/2001 - 01.03.2001
IMGC 03/2001 19.03.2001 15.05.2001
INMETRO 06/2001 31.05.2001 30.08.2001
NIST 08/2001 19.08.2001 15.09.2001
AIST 09/2001 10.09.2001 26.02.2003
NIM 10/2001 05.02.2002 07.03.2002
KRISS 11/2001 30.03.2002 07.11.2002
CMI 12/2001 22.05.2002 05.09.2002
GUM 01/2002 17.09.2002 25.10.2002
NPL 02/2002 20.10.2002 16.01./25.02.2003
PTB 03/2002 20.02.2003 18.03.2003
Moreover, because not all participants were able to measure the three hardness
scales HV0,2, HV1 and HV30, which is justified by different economical needs, it was
allowed to participate only in a part of the scheduled measurement programme.
Accordingly, Table 3 gives an overview over the measurements carried out by the
participants.
5
Table 3: Overview over the measurements carried out by the participants
Laboratory HV0,2 HV1 HV30 Ref. Ref. Ref.

indent indent indent
HV0,2 HV1 HV30
PTB x x x x X x
IMGC x x x x X x
INMETRO x x
NIST x x x X
AIST x x x x X x
NIM x x x x X x
KRISS x x x x X x
CMI x x x x X x
GUM x x x x X x
NPL x x x x X x
PTB x x x x X x
3 Standards
3.1 Description
In the key comparison three sets of hardness reference blocks for the Vickers
hardness scales HV0,2, HV1 and HV30 consisting each of three hardness reference
blocks with the hardness levels 240 HV, 540 HV and 840 HV (that is altogether nine
blocks) are used. The dimensions are length 60 mm, width 60 mm, thickness 10 mm.
The upper side of the blocks which is the measurement surface is finished. The
blocks are manufactured as commercial products by Buderus Co., Germany. For the
comparison on the hardness reference blocks on the measurement surface was
engraved a grid with 13 x 13 = 169 fields. At the left and the lower edge of the blocks
are engraved numbers from 1 to 13 in order to define coordinates of the fields. The
direction along the lower edge of the blocks defines the X-direction (lines),
correspondingly the direction along the left edge of the blocks defines the Y-direction
(rows) (see Fig. 1).
13
12
11
10
9
Y
8
7
6
5
4
3
2
1
1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 1: Layout of the grid on the measurement surface of the hardness reference
blocks
6
3.2 Handling
It is recommended to clean the blocks after unpacking with alcohol and then sign all
fields reserved for your institute with a fiber pen on the left top corner. After
measurement all dots on the blocks must be removed before packing in order to
avoid corrosion.
4 Measurand
The measurands used in this comparison were of two kinds. Firstly, the mean value
each of eight hardness measurements on a hardness reference block had to be
determined. The hardness measurements were made in the hardness scales HV0,2,
HV1 and HV30 each for the nominal hardness levels 240 HV, 540 HV and 840 HV.
The procedure of the hardness measurement is defined in ISO 6507-1 and -3.
Secondly, the mean diagonal length of nine reference indents had to be determined
according to ISO 6507-1 and -3. The reference indents represent indents for the
hardness levels 240 HV, 540 HV and 840 HV, each for the Vickers scales HV0,2,
HV1 and HV30.
5 Methods of measurement
The methods of measurement and the measuring devices used by the participants
are described in Appendix A1.
6 Stability of the standards
In order to evaluate the stability of the standards, the hardness reference blocks, the
pilot laboratory has carried out measurement in the beginning and in the end of the
comparison. The results are summarised in Table 4.
Table 4: Measurement results in the beginning and in the end of the comparison by
the pilot laboratory
Measurand Result in the Result in Diff. 1-2 Meas. l1-2l/ U
begin. (1) the end (2) Uncert. U
240 HV0,2, HV 248,19 252,27 -4,08 5,03 0,81
540 HV0,2, HV 528,42 540,63 -12,21 16,25 0,75
840 HV0,2, HV 839,78 858,07 -18,29 31,33 0,58
240 HV1, HV 244,06 243,15 +0,91 2,18 0,42
540 HV1, HV 539,03 537,76 +1,27 7,18 0,18
840 HV1, HV 831,49 835,79 -4,30 13,86 0,31
240 HV30, HV - 238,00 - - -
540 HV30, HV - 522,89 - - -
840 HV30, HV - 817,86 - - -
r.i.240 HV0,2, m 38,52 38,36 +0,16 0,2 0,80
r.i.540 HV0,2, m 26,38 26,01 +0,37 0,2 1,85
r.i.840 HV0,2, m 21,08 20,90 +0,18 0,2 0,90
r.i.240 HV1, m 87,55 87,41 +0,14 0,6 0,23
r.i.540 HV1, m 58,58 58,59 -0,01 0,4 0,03
r.i.840 HV1, m 47,38 47,20 +0,18 0,3 0,60
r.i.240 HV30, m 482,48 482,47 +0,01 1,1 0,01
r.i.540 HV30, m 326,03 326,65 -0,62 1,1 0,56
r.i.840 HV30, m 259,35 258,88 +0,47 1,1 0,43
7
In the last row the difference between first and second measurement 1-2 is
compared with the measurement uncertainty. If the difference l1-2l/U > 1, it means,
that the difference 1-2 cannot be explained by the uncertainty but can be traced back
to any change of the hardness reference blocks during the period of the comparison.
The exceeding at the reference indent for 540 HV0,2 is insignificant.
Therefore, one can conclude that the used hardness reference blocks remained
stable.
Although the conclusion about the stability according to the difference between the
first and the last measurement seems to be justified, the analysis of the
measurement results in dependence on the time of the comparison revealed, that in
the course of the measurements eight of the nine used hardness reference blocks
showed like a quadratic function first a tendency to lower hardness values and then
the hardness values rose approximately to the original level.
Fig. 2 shows this change on the example of the hardness reference block 240 HV1.
Hardness measurement deviation HV, HV
4
240HV1
-2
-4
0 5 10 15 20 25
Time t, months
Fig. 2: Apparent hardness change of the hardness reference block 240 HV1 over
the time of the comparison
From the fitted quadratic function a maximum hardness change of 8,9 HV = 3,7 %
was derived. An overview of the change of all hardness reference blocks is given in
Table 5.
8
Table 5:
Evaluation of the apparent hardness change of the used hardness reference blocks
over the time of the comparison
Hardness reference Hardness change, HV Relative hardness change,

block %
240HV0,2 8,9 3,7
240HV1 5,4 2,3
240HV30 -- --
540HV0,2 6,8 1,3
540HV1 8,3 1,5
540HV30 4,7 0,9
840HV0,2 32,8 3,9
840HV1 14,7 1,8
840HV30 6,3 0,8
The apparent hardness changes shown in Table 5 differ with the test force. For the
scales with smaller test forces (HV0,2 and HV1) the hardness changes are larger
than in the case of the hardness scale HV30. This means that the found quadratic
course of the hardness change over the time only to a very small part can be
attributed to a real hardness change. Also the mechanism of a reversible hardness
change in the hardness reference blocks is not clear. Because the transport box of
the blocks did not contain monitoring sensors for temperature and other environment
influences, afterwards we can not determine whether intolerable temperature shocks
and other events having negative effects happened during the transport of the
blocks.
Therefore, the above conclusion about the stability of the hardness reference blocks
can be maintained.
7 Measurement results
In the following tables 6 to 8 the results for the hardness reference blocks with
hardness levels of 240 HV, 540 HV and 840 HV are summarised. The results are
expressed by the mean values, the standard deviations s8 of each 8 repetition
measurements and the standard deviations between the institutes sInst.
9
Table 6:
Results of the measurements for the hardness reference blocks with hardness level
240 HV
HV0,2 HV1 HV30

Institute Mean value Std.dev. Mean value Std. dev. Mean value .Std. dev.
PTB 248,19 2,67 244,06 1,79
IMGC 254,30 4,34 241,11 3,18 237,83 0,88
INMETRO 244,83 0,77
NIST 244,88 3,31 236,13 1,81
NMIJ 241,59 3,92 238,41 1,71 236,11 1,21
NIM 242,36 2,77 237,88 0,86 236,40 0,44
KRISS 247,58 3,48 238,83 1,88 241,96 0,46
CMI 245,21 3,00 240,73 0,66 238,73 0,39
GUM 242,14 2,56 238,56 2,28 242,58 0,77
NPL 253,65 2,92 240,33 3,55 238,55 0,53
PTB 252,27 2,99 243,15 2,46 238,00 0,56
Mean value 247,20 239,92 239,44

Std.dev. 4,83 2,44 2,99
Table 7:
540 HV
HV0,2 HV1 HV30

Institute Mean val.x8 Std.dev.s8 Mean val.x8 Std. dev.s8 Mean val.x8 Std. dev.s8
PTB 528,42 5,81 539,03 5,97
IMGC 548,16 6,12 544,84 5,71 525,45 1,35
INMETRO 534,94 2,08
NIST 528,50 3,34 522,50 1,77
NMIJ 529,34 6,24 524,45 5,71 516,68 1,90
NIM 536,63 1,87 528,56 2,10 518,88 0,79
KRISS 538,79 4,61 533,89 2,94 528,41 1,57
CMI 522,83 5,50 535,38 1,30 520,73 0,57
GUM 529,76 6,11 535,55 2,72 527,90 1,30
NPL 547,36 13,07 536,63 4,34 524,86 1,06
PTB 540,63 5,22 537,76 1,59 522,89 2,50
M.val.xInst 535,04 533,86 524,53

Std.dev.sInst 8,61 6,84 5,55
10
Table 8:
840 HV
Results for 840 HV hardness reference blocks
HV0,2 HV1 HV30
Institute Mean val.x8 Std.dev. s8 Mean val.x8 Std. dev. s8 Mean val.x8 Std. dev. s8
PTB 839,78 9,38 831,49 6,14
IMGC 864,20 44,79 875,40 10,05 820,85 2,15
INMETRO 841,55 1,43
NIST 827,00 7,09 815,38 6,35
NMIJ 832,18 13,40 825,54 7,39 811,59 5,93
NIM 833,96 5,00 828,80 3,24 815,91 1,71
KRISS 840,25 3,77 839,75 5,77 831,86 1,68
CMI 813,37 4,02 826,82 4,45 810,64 1,08
GUM 830,00 5,77 831,77 3,26 818,03 2,20
NPL 897,57 13,60 839,32 7,54 837,17 1,55
PTB 858,07 10,22 835,79 6,62 817,86 3,39
M.val.xInst 843,64 835,01 822,83

Std.dev. sInst 23,97 15,91 11,25
Fig. 3 presents as an extract from tables 6 to 8 the reproducibility sInst over the
hardness in absolute hardness values and Fig. 4 the same relationship in relative
values of reproducibility.
HV0,2
25
HV1
HV30
20
Reproducibility sInst, HV
15
10
0
200 300 400 500 600 700 800 900
Vickers hardness, HV
Fig. 3: Reproducibility sInst over Vickers hardness for the three scales HV0,2, HV1
and HV30
11
3,0 HV0,2
HV1
HV30
Relative reproducibility sInstrel, %
2,5
2,0
1,5
1,0
200 300 400 500 600 700 800 900
Fig.4: Relative reproducibility sInstrel over Vickers hardness for the three scales
HV0,2, HV1 and HV30 (fitted curves)
The Fig. 3 and 4 illustrate the presently achievable degree of agreement for Vickers
measurements in the scales HV0,2, HV1 and HV30 carried out by the participating
national metrology institutes. Whereas for the high hardness level of 840 HV the
uncertainty (expressed by the reproducibility) increases with decreasing test force,
for the low hardness level of 240 HV the influence from the diagonal measurements
becomes more evident. This may be the reason that sInstrel at 240 HV for HV30 is
higher than for HV1.
8 Analysis
8.1 Reference values on the basis of the uncertainty determined by the
participants
For the calculation of the reference values two ways were tried:
1) Reference values based on the arithmetic mean
2) Reference values based on the weighted mean
In the following these two evaluation methods are compared with each other.
8.1.1 Reference values based on the arithmetic mean

The arithmetic mean value is calculated as follows:
1 n
x = xi (1)
n i =1
with
x1, x2,...,xn = measurement results of n participants
n = number of participants
The uncertainty of the arithmetic mean value is expressed by the confidence interval:
ts
ux = (2)
n
12
with
t = factor of Student distribution (t = f(,n))
s = standard deviation of the measurement results of n participants
Finally, the coefficient En, which evaluates the agreement between the measurement
deviations found in the comparison and the uncertainties stated by the participants, is
calculated with the following expression:
xlab x ref
En = (3)
2
U lab + U ref
2
where xlab is the measurement result of the participating laboratory, xref is the
reference value calculated by the weighted mean value, Ulab is the uncertainty stated
by the participating laboratory and Uref is the uncertainty of xref calculated by the
uncertainty of the weighted mean value.
For the reference values only the measurement results with En 1 are considered.
Using the uncertainties in Appendix A2 one gets the following references values.
Table 9:
Reference values and their uncertainties for the hardness measurements (evaluation
by the participants)
Hardness scale Ref. value xref Uncertainty of reference value Uref

Absolute unc., HV Relative unc., %
240 HV0,2 247,09 4,57 1,85
240 HV1 239,46 1,70 0,71
240 HV30 239,44 2,44 1,02
540 HV0,2 535,78 7,17 1,34
540 HV1 533,28 6,25 1,17
540 HV30 524,53 4,52 0,86
840 HV0,2 844,07 20,69 2,45
840 HV1 835,40 13,71 1,64
840 HV30 822,83 9,17 1,11
8.1.2 Reference values based on the weighted mean

The reference values are calculated as the weighted mean based on the
uncertainties of the participants. The weighted mean is calculated as follows:
x u 2 + x u 2 + ... + x n u n2
x w = 1 1 2 2 22 (4)
u1 + u 2 + ... + u n 2
with
x1, x2,...,xn = measurement results of n participants
u1, u2,...,un = measurement uncertainties of n participants
The uncertainty of the weighted mean according to the law of error propagation
follows from:
U xw = (u1 S ) 2 + (u 2 S ) 2 + ... + (u n S ) 2 (5)

with
S = u12 + u 22 + ... + u n2
13
Using the uncertainties in Table 14 (see paragraph 8.2) one gets the following
references values.
Table 10:
by the participants)
Hardness scale Ref. value xref Uncertainty of reference value Uref

Absolute unc., HV Relative unc., %
240 HV0,2 246,00 2,54 0,45
240 HV1 238,48 1,37 0,42
240 HV30 238,58 0,54 0,23
540 HV0,2 531,29 6,65 1,13
540 HV1 534,89 4,36 0,82
540 HV30 524,97 1,39 0,26
840 HV0,2 827,76 11,86 1,36
840 HV1 828,36 6,86 0,83
840 HV30 813,48 3,04 0,37
In the case of 840 HV30 the measurement results showed a distribution with two
peaks. Because 6 values where distributed around 815 HV and 3 values around 835
HV, the reference value was taken from the majority of the measurement results.
Fig. 5 depicts the dependency of the uncertainty on the hardness level and the test
force (resp. hardness scale).
HV0,2
Relative uncertainty of reference values, Uref, %
1,4
HV1
HV30
1,2
1,0
0,8
0,6
0,4
0,2
200 300 400 500 600 700 800 900
Fig. 5: Dependency of the relative uncertainty of the reference values on the

hardness level and the test force (resp. hardness scale)
Fig. 5 clarifies that the relative uncertainty of the reference values as expected
increases with the hardness, whereby smaller test forces deliver higher uncertainties
than larger test forces. This reflects the fact that the Vickers hardness uncertainty is
strongly influenced by the uncertainty of the diagonal measurement (see also ch.
8.4).
It should be noted that the uncertainties of the reference values are obviously the
smallest uncertainties reached in the field of Vickers measurements worldwide so far.
These uncertainties can be interpreted as the present limits of Vickers
14
measurements in the investigated range of hardness scales. This is one important
outcome of this Vickers key comparison.
In the following Fig. 6 to 14 the measurement results are shown in connection with
the reference values and the uncertainties stated by the participants.
270 240 HV0,2

265 Reference value: 246,00 HV 0,2
260
255
250
245
240
235
230
225
RO
--
GC
L
IJ
B
ST
I
CM
NP
PT
NM
IS
GU
NI
NI
ET
IM
KR
M
IN
Fig. 6: Measurement results for the hardness scale 240 HV0,2 related to the
reference value
240 HV1
Reference value: 238,48 HV1
245
240
235
230
--
S
L
RO
IJ
GC
B
M
I
ST
CM
NP
PT
IS
NM
GU
NI
NI
IM
ET
KR
M
IN
Fig. 7: Measurement results for the hardness scale 240 HV1 related to the reference
value
15
250 240 HV30
248
246
244
242
240
238
236
234
232
230
--
RO
IJ
L
GC
B
ST
I
CM
NP
PT
NM
IS
GU
NI
NI
ET
IM
KR
M
IN
Fig. 8: Measurement results for the hardness scale 240 HV30 related to the
reference value
580 540 HV0,2

Reference value: 531,29 HV0,2
560
540
520
500
480
--
S
L
RO
IJ
GC
B
M
I
ST
CM
NP
PT
IS
NM
GU
NI
NI
IM
ET
KR
M
IN
reference value
16
560 540 HV1
555 Reference value: 534,89 HV1
550
545
540
535
530
525
520
515
510
505
500
495
--
M
L
RO
IJ
GC
S
M
B
ST
CM
NP
PT
NM
IS
GU
NI
NI
IM
ET
KR
M
IN
Fig. 10:Measurement results for the hardness scale 540 HV1 related to the reference
value
540 HV30
545 Reference value: 524,97 HV30
540
535
530
525
520
515
510
505
--
RO
L
IJ
GC
B
M
I
ST
CM
NP
PT
IS
NM
GU
NI
NI
ET
IM
KR
M
IN
reference value
17
940 840 HV0,2
920 Reference value: 827,76 HV0,2
900
880
860
840
820
800
780
760
740
M
RO
IJ
--
L
GC
B
ST
I
CM
NP
PT
NM
IS
GU
NI
NI
ET
IM
KR
M
IN
reference value
910
840 HV1
900
890
880
870
860
850
840
830
820
810
800
790
780
770
--
RO
S
M
L
IJ
C
B
I
ST
CM
NP
PT
IS
NM
G
GU
NI
NI
IM
ET
KR
M
IN
reference value
18
860
840 HV30
855
850
845
840
835
830
825
820
815
810
805
800
795
790
--
RO
L
IJ
GC
B
M
I
ST
CM
NP
PT
IS
NM
GU
NI
NI
ET
IM
KR
M
IN
reference value
The question whether the measurement deviations found in the comparison can be
explained by the uncertainties stated by the participants is evaluated with the
coefficient En. In the following Tables 11 and 12 the coefficients En are summarised
for all participants and hardness scales, first for reference values based on the
arithmetic mean and then based on the weighted mean.
Table 11:
Coefficients En for all participants and all hardness scales (based on arithmetic
mean)
En values
240 HV 540 HV 840 HV
240HV0,2 240HV1 240HV30 540HV0,2 540HV1 540HV30 840HV0,2 840HV1 840HV30
IMGC 0,67 0,37 0,50 0,39 0,76 0,14 0,31 1,28 0,16
INMETRO 1,06 0,99 1,05
NIST 0,33 1,09 0,46 1,12 0,48 1,05
NMIJ 0,37 0,15 1,06 0,14 0,39 1,04 0,13 0,22 0,78
NIM 0,54 0,42 1,03 0,04 0,37 0,82 0,19 0,26 0,49
KRISS 0,06 0,17 0,58 0,13 0,06 0,52 0,08 0,20 0,64
CMI 0,25 0,16 0,26 0,90 0,11 0,73 1,11 0,29 1.22
GUM 0,55 0,21 0,56 0,27 0,21 0,28 0,34 0,19 0,25
NPL 0,35 0,07 1,51
PTB 0,57 0,70 0,51 0,24 0,35 0,26 0,33 0,02 0,39
Mean 0,42 0,41 0,66 0,32 0,42 0,54 0,37 0,44 0,72
value
19
Table 12:
Coefficients En for all participants and all hardness scales (based on weighted mean)
En values
240 HV 540 HV 840 HV
IMGC 0,84 0,61 0,22 0,53 0,92 0,35 0,58 1,54 0,79
INMETRO 1,45 1,18 0,89
NIST 0,22 0,81 0,18 1,11 0,02 1,06
NMIJ 0,31 0,01 1,07 0,04 0,32 1,14 0,05 0,13 1,45
NIM 0,47 0,17 1,11 0,24 0,27 0,89 0,12 0,10 1,10
KRISS 0,22 0,10 1,00 0,33 0,23 0,80 0,28 0,47 0,37
CMI 0,13 0,29 0,33 0,60 0,21 1,04 0,66 0,15 3,87
GUM 0,49 0,02 0,85 0,07 0,40 0,39 0,06 0,05 0,57
NPL 0,28 0,62 2,98
PTB 0,78 0,91 0,20 0,47 0,53 0,16 0,79 0,25 1,10
Mean 0,42 0,37 0,72 0,31 0,50 0,73 0,32 0,47 1,46
value
As the mean values of En for the hardness scales express the degree of difficulty to
measure in the corresponding hardness scales, Tables 11 and 12 deliver the
somewhat astonishing result that the uncertainties with increasing test force (resp.
with increasing diagonal length) in the mean are underestimated. On the other side,
the uncertainties for HV0,2 as compared with the uncertainties for HV30 generally
are estimated more realistically. This result is also confirmed by the investigation of
the influence of the diagonal measurement uncertainty.
The comparison of the En values derived from the arithmetic and from the weighted
mean yield the result that the En values based on the arithmetic mean deliver less
exceedings of the limit En = 1 than the En values based on the weighted mean. This
clarifies that the weigthed mean is not well suited for a small number of
participants (n = 10), because under the point of view of statistics it is difficult to
declare the results of 4 or 5 participants out of 10 to outliers.
8.2 Reference values on the basis of a unified estimation of the hardness

measurement uncertainty
The application of the EA draft "Guideline to the estimation of the uncertainty of the
Brinell and the Vickers measuring method" already guaranteed a rather uniform
approach of all participants to the estimation of the measurement uncertainty.
Nevertheless, the amount of considered influence quantities was different, and the
type of distribution was also not always considered. So the individual application of
this guideline in several cases led to uncertainty values which are not sufficiently
comparable.
Therefore a unified estimation of the measurement uncertainty was made on the
following basis.
The calculation scheme can be seen from the following example.
20
Table 13:
Calculation scheme for the unified estimation of the measurement uncertainty
Influencing quantity X i Symbol Unit Value xi si ai u 2 (x i ) ci u 2 (y i ) i u i 4 (y)/ i

Test force F F N 1,96 0,000 9,80E-03 3,2E-05 1,3E+02 0,0E+00 5,1E-01 8 3,3E-02
Indentation diagonal length d mm 0,039 0 0,0004 5,3E-08 -1,3E+04 0,0E+00 8,7E+00 9 8,5E+00
Plane angle
d 136 -8,7E-04 0,00175 1,0E-06 5,0E+01 -4,4E-02 2,5E-03 10 6,4E-07
Tip radius r r mm 3,0E-04 1,0E-04 3,3E-09 -7,1E+03 -2,1E+00 1,7E-01 10 2,8E-03
Length of line of junction c c mm 4,0E-04 1,0E-04 3,3E-09 9,0E+03 3,6E+00 2,7E-01 10 7,1E-03
Total 1,41 9,7E+00 8,5E+00
Combined standard uncertainty u (H) 3,1E+00 eff 11
Confidence level 95%
Coverage factor 2,2
Expanded standard uncertainty U (H) 6,8 HV
Expanded standard uncertainty U(H)+I HI 8,3 HV
Relative Expanded standard uncertainty U rel (H) 3,3 %
Hardness 247,5 HV
From the influencing quantities Xi measurement deviations xi and uncertainties in

the form of standard deviation si (type A) and ai (type B) are considered.
Standard uncertainty:
a2
u 2 ( xi ) = s i2 i (6)
3
Sensitivity coefficients:
HV 0,204 sin / 2
c1 = =
F d2
HV 0,408 F sin / 2
c2 = =
d d3
HV 0,102 F cos / 2
c3 = = (7)
d2
HV 0,204 F sin / 2 r
c4 = = (1,099 + 1,1515 )
r d 3
d
HV 0,2856 F sin / 2
c5 = =
c d3
Single hardness deviation:
H i = ci xi (8)
Variances:
u 2 ( y i ) = ci2 u 2 ( xi ) (9)
Combined standard uncertainty:
n
u( H ) = u
i =1
2
( yi ) (10)
Sum of hardness deviations:

n
H = H i (11)
i =1
Effective degrees of freedom, according to the Welch-Satterthwaite formula:

u 4 ( y)
eff = 4 (12)
ui ( y)

i =1 i
Coverage factor:
k = f (v eff , P) (13)
Expanded standard uncertainty:
21
U ( H ) = k u ( H ) + H (14)
Relative expanded standard uncertainty:
U (H )
U rel ( H ) = 100, 0 0 (15)
H
According to this unified procedure for the estimation of measurement uncertainty,
the following measurement uncertainties for the participants were received. If the
participants had omitted the indenter tip radius because they did not have the facility
to measure it, a value of r = 0,5 m with an uncertainty of 0,2 m was set in.
Table 14:
Results of the relative measurement uncertainty (in %) according to the estimation of
the participant (part) and to the unified procedure (unif)
Institute 240HV0,2 240HV1 240HV30 540HV0,2 540HV1 540HV30 840HV0,2 840HV1 840HV30
part unif part unif part unif part unif part unif part unif part unif part unif part unif
IMGC 3,9 4,9 1,7 2,0 0,9 0,4 5,7 7,2 2,5 3,2 1,0 0,6 7,1 9,0 3,2 4,0 1,0 0,7
INMETRO 1,8 3,3 1,8 2,2 1,8 2,7
NIST 2,0 4,4 1,1 2,0 2,7 6,5 1,4 3,0 3,5 8,2 1,6 3,6
NMIJ 5,8 4,9 2,9 2,5 0,9 0,7 8,9 7,3 4,2 3,7 1,2 1,1 11,1 9,2 5,2 4,6 1,4 1,3
NIM 3,1 5,6 1,4 2,5 0,7 0,7 3,9 8,5 2,1 3,8 1,0 1,1 5,9 10,6 2,6 4,7 1,3 1,3
KRISS 2,9 5,3 1,4 2,4 1,5 0,9 4,0 8,1 1,6 3,2 1,1 0,7 5,0 9,5 2,1 4,6 1,3 1,1
CMI 2,4 6,4 3,2 2,7 0,5 0,9 2,3 9,3 3,2 4,1 0,5 1,1 2,2 11,8 3,2 5,1 0,5 1,3
GUM 3,3 5,0 1,7 3,1 2,1 2,5 3,9 7,5 1,7 3,8 2,0 2,5 4,3 9,3 1,7 3,8 2,1 2,5
NPL 0,3 0,6 0,3 0,9 0,3 1,1
PTB 3,1 5,1 2,0 3,1 0,6 0,9 3,4 5,6 2,0 3,3 0,9 1,3 4,3 6,9 2,2 3,6 1,1 1,6
Mean value 3,3 5,2 1,9 2,5 1,0 1,2 4,4 7,5 2,3 3,5 1,1 1,3 5,4 9,3 2,7 4,3 1,2 1,5
Variation 3,8 2 2,1 1,1 1,8 2,9 6,6 3,7 2,8 1,1 1,7 1,9 8,9 4,9 3,6 1,5 1,8 2
Table 14 elucidates that the application of the unified procedure for the measurement
uncertainty in most cases led to a reduction of the variation (difference between
maximum and minimum value of the uncertainty). An increase of the variation is
observed for the scale HV30 which is due to a relatively big change of uncertainty in
one single case. On the other hand, the level of uncertainty, as expressed by the
mean value, increased. The reason is that the hardness deviations due to the
indenter calibration are rather large. This fact underlines that it is necessary to
improve the calibration methods of the indenter.
In Tables 15 and 16 the coefficients En for all participants and all hardness scales
according to the evaluation by the unified procedure for the estimation of the
measurement uncertainty and the corresponding reference values are summarised,
first for the weighted mean and then for the arithmetic mean.
22
Table15:
Coefficients En in the case of the unified procedure for the estimation of the
measurement uncertainty (based on arithmetic mean)
En values
240 HV 540 HV 840 HV
HV0,2 HV1 HV30 HV0,2 HV1 HV30 HV0,2 HV1 HV30
IMGC 0,55 0,32 0,62 0,31 0,63 0,17 0,26 1,06 0,18
INMETRO 0,66 0,84 0,76
NIST 0,19 0,67 0,20 0,63 0,24 0,62
NMIJ 0,44 0,17 1,12 0,16 0,44 1,10 0,15 0,24 0,79
NIM 0,33 0,25 1,02 0,02 0,22 0,77 0,11 0,16 0,48
KRISS 0,04 0,11 0,78 0,07 0,03 0,68 0,05 0,11 0,70
CMI 0,12 0,19 0,22 0,26 0,09 0,51 0,31 0,19 0,85
GUM 0,39 0,12 0,48 0,15 0,11 0,24 0,17 0,11 0,22
NPL 0,31 0,05 1,12
PTB 0,39 0,48 0,45 0,16 0,24 0,20 0,23 0,01 0,31
Mean 0,31 0,29 0,63 0,17 0,30 0,51 0,19 0,31 0,60
Table16:
measurement uncertainty (based on weighted mean)
En values
240 HV 540 HV 840 HV
IMGC 0,61 0,39 0,20 0,31 0,65 0,30 0,30 1,10 0,25
INMETRO 0,87 0,90 0,83
NIST 0,12 0,57 0,17 0,59 0,18 0,60
NMIJ 0,37 0,10 1,08 0,14 0,41 1,35 0,10 0,22 0,96
NIM 0,27 0,18 0,92 0,03 0,20 0,92 0,07 0,13 0,56
KRISS 0,10 0,04 1,80 0,08 0,06 1,03 0,00 0,13 0,97
CMI 0,06 0,24 0,32 0,24 0,12 0,60 0,27 0,17 1,03
GUM 0,32 0,06 0,75 0,12 0,13 0,26 0,12 0,08 0,22
NPL 0,32 0,09 1,53
PTB 0,45 0,53 0,02 0,17 0,26 0,22 0,28 0,05 0,34
Mean 0,29 0,26 0,70 0,16 0,30 0,63 0,17 0,31 0,74
Compared with the uncertainty estimation by the participants (Tables 11 and 12) the
En values in Table 15 and 16 are a little bit smaller. This is due to the harmonisation
of the uncertainty estimation by the unified procedure. The values in Table 15 exhibit
the same tendency as those in Table 11 that the En values are increasing with the
test force as can be seen by the mean values. This means that the uncertainty for the
largest test force (and the largest indentation size) in several cases is
underestimated.
The comparison of the En values according to the unified method based on the
weighted mean and the arithmetic mean shows clearly that the use of the arithmetic
mean delivers the smallest number of exceedings of the limit En = 1 (5 exceedings for
the arithmetic mean against 7 exceedings for the weighted mean). Obviously, the use
of the weighted mean leads to an overestimation of single results if the number of the
participants is small (n = 10 in the case of this key comparison). Therefore the use of
23
the arithmetic mean for the calculation of the reference values and the En values
seems to be most suited in order to evaluate this key comparison.
On the basis of the measurement uncertainties according to the unified procedure
based on the arithmetic mean the reference values were recalculated. Measurement
results with En 1 have been excluded. The calculation of the reference values
according to the estimation of the participants and to the unified method is given in
Appendix A2. The results are shown in the following Table 17.
Table 17:
by the unified procedure based on the arithmetic mean, values with En 1)
Hardness Reference value xref, Uncertainty of reference value

scale HV Uref
Abs. unc., HV Rel. unc., %
240 HV0,2 246,28 4,57 1,86
240 HV1 239,81 2,08 0,87
240 HV30 239,15 1,97 0,82
540 HV0,2 534,33 8,80 1,65
540 HV1 534,35 6,10 1,14
540 HV30 525,51 5,03 0,96
840 HV0,2 840,81 25,38 3,02
840 HV1 831,41 16,82 2,02
840 HV30 819,35 11,25 1,37
The following Fig. 15 until 23 show the measurement results under consideration of
the uncertainty according to the unified procedure.
25
240 HV0,2
20 reference value 246,28 HV0,2
Deviation from reference value, HV
15
10
-5
-10
-15
-20
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
reference value (unified procedure)
24
15 240 HV1
Reference value 239,81 HV1
10
-5
-10
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
14 240 HV30
12
10
8
6
4
2
0
-2
-4
-6
M
UM
RO
GC
B
ST
--
L
I
IJ
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
25
540 HV0,2
60
Reference value 534,33 HV0,2
40
20
-20
-40
-60
-80
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
35
30 540 HV1
25 Reference value 534,35 HV1
20
15
10
5
0
-5
-10
-15
-20
-25
-30
-35
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
26
25
540 HV30
15
10
-5
-10
-15
M
--
B
L
M
RO
GC
I
ST
IJ
CM
NP
PT
IS
NM
NI
GU
NI
IM
ET
KR
M
IN
120 840 HV0,2

100 Reference value 840,81 HV0,2
80
60
40
20
0
-20
-40
-60
-80
-100
-120
-140
M
M
RO
B
ST
--
L
I
IJ
GC
S
CM
NP
PT
IS
NM
GU
NI
NI
IM
ET
KR
M
IN
27
90
80
840 HV1
70
60
50
40
30
20
10
0
-10
-20
-30
-40
-50
-60
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
45 840 HV30
35
30
25
20
15
10
5
0
-5
-10
-15
-20
-25
-30
M
UM
RO
--
IJ
GC
B
ST
L
I
CM
NP
PT
IS
NM
NI
NI
ET
IM
G
KR
M
IN
8.3 Influence of the inhomogeneity of hardness of the hardness reference

blocks
The inhomogeneity of the hardness distribution on the hardness reference blocks is
evaluated on the following basis. According to the technical protocol each participant
had to carry out eight measurements on each hardness reference block. The
distribution of these eight measurement locations on the block was such that two
consecutive locations were rather near to each other so that one can assume that the
hardness change between these two locations is sufficiently small as compared with
the hardness change due to the repeatability of the hardness standard machine.
Therefore, the standard deviation based on the differences between the four groups
of each two neighbouring locations can be used to express the inhomogeneity of
hardness distribution on the hardness reference blocks.
28
If the measuring values on the eight locations are denominated as:
x1, x2, ... x8,
then the differences of the four groups of neighbouring locations are:
1 = (x1 - x2)/2
2 = (x3 - x4)/2
3 = (x5 - x6)/2
4 = (x7 - x8)/2 (16)
Further, the standard deviation due to inhomogeneity of hardness is expressed by:
1 2
s Inh = (1 + 22 + 23 + 24 ) (17)
3
Table 18 summarises the results of the inhomogeneity of the hardness reference

blocks, as derived from the measuring results of CMI. The values of this institute
were chosen because the repeatability of their results is sufficiently small. Therefore
one can suppose that the influences from the inhomogeneity and the repeatability are
rather well separated.
Table 18:
Results of the inhomogeneity of the hardness reference blocks, as derived from the
measuring results of CMI
Standard deviation due to inhomogeneity s Inh

Related to diagonal Related to Relative inhomo-
length d hardness HV geneity s Inh /HV , %
240HV0,2 0,14 0,83 0,35
240HV1 0,12 0,64 0,27
240HV30 0,27 0,27 0,11
540HV0,2 0,10 4,01 0,74
540HV1 0,02 0,44 0,08
540HV30 0,17 0,54 0,10
840HV0,2 0,05 3,33 0,40
840HV1 0,12 4,11 0,49
Table 18 shows that the relative inhomogeneity depends less on the hardness level,
but more on the hardness scale. The relative inhomogeneity increases with
decreasing test force resp. with decreasing indent size. This fact clarifies that the
inhomogeneity is still superposed by the repeatability of the length measurement. As
a rough estimate one can conclude that the relative inhomogeneity of hardness
distribution on the used hardness reference blocks is not larger than 0,5 %.
8.4 Influence of the length measuring deviation

Because at Vickers measurements the length measuring deviation significantly
influences the uncertainty of measurements, in the technical protocol the
measurement of reference indents was included.
29
In the Tables 19 to 21 the results for the measurements of the reference indents are
summarised. At this the repeatability of the measurements is expressed by sr and
calculated as follows:
s + sv
sr = h (18)
2
with
sh = standard deviation of the diagonal measurements in horizontal direction
sv = standard deviation of the diagonal measurements in vertical direction
The standard deviation sInst represents the reproducibility of measurements between

the participating laboratories.
Table 19:
Results for 240 HV reference indents
HV0,2 HV1 HV30

Institute Mean value Std.dev. sr Mean value Std. dev. sr Mean value . Std. dev. sr
PTB 38,52 0,09 87,55 0,05 482,48 0,68
IMGC 38,19 0,12 87,70 0,16 483,66 0,62
INMETRO 479,40 0,7
NIST 38,90 0,09 87,90 0,05
NMIJ 39,18 0,03 88,19 0,06 485,39 0,15
NIM 39,19 0,05 88,29 0,08 485,95 0,14
KRISS 38,65 0,41 88,65 0,09 480,89 0,00
CMI 38,96 0,13 88,01 0,17 483,96 0,58
GUM 38,95 0,06 87,92 0,03 482,55 0,06
NPL 38,17 0,11 87,27 0,23 484,37 0,02
PTB 38,36 0,16 87,41 0,23 482,47 0,88
Mean value 38,71 0,13 87,89 0,12 483,11 0,38

Std.dev. sInst 0,38 0,42 1,99
Table 20:
HV0,2 HV1 HV30
Institute Mean value Std.dev. sr Mean value Std. dev. Sr Mean value . Std. dev. sr
PTB 26,38 0,14 58,58 0,12 326,03 0,82
IMGC 26,01 0,15 58,31 0,12 325,39 0,35
INMETRO 321,80 0,22
NIST 26,30 0,06 58,80 0,06
NMIJ 26,47 0,03 59,47 0,03 328,14 0,00
NIM 26,24 0,08 59,25 0,09 327,85 0,09
KRISS 26,30 0,07 58,65 0,00 323,30 0,00
CMI 26,61 0,06 59,24 0,20 326,30 0,48
GUM 26,25 0,05 59,07 0,03 325,47 0,08
NPL 25,43 0,13 58,34 0,09 325,87 0,01
PTB 26,01 0,09 58,59 0,13 326,65 0,28
Mean value 26,20 0,09 58,83 0,09 325,68 0,23

Std.dev. sInst 0,33 0,41 1,92
30
Table 21:
HV0,2 HV1 HV30

Institute Mean value Std.dev. sr Mean value Std. dev. sr Mean value . Std. dev. sr
PTB 21,08 0,08 47,38 0,11 259,35 0,22
IMGC 20,74 0,14 46,03 0,12 260,34 0,26
INMETRO 255,50 0,56
NIST 21,10 0,03 47,50 0,03
NMIJ 21,11 0,00 47,40 0,06 261,81 0,03
NIM 21,23 0,09 47,33 0,07 261,30 0,10
KRISS 21,01 0,11 47,12 0,08 257,85 0,00
CMI 21,21 0,09 47,33 0,20 260,54 0,40
GUM 21,00 0,06 47,42 0,91 259,40 0,09
NPL 20,51 0,29 46,87 0,16 257,79 0,02
PTB 20,90 0,21 47,20 0,10 258,88 0,33
Mean value 20,99 0,11 47,16 0,18 259,28 0,20

Std.dev. sInst 0,22 0,44 1,88
Fig. 24 shows the relative reproducibility of the diagonal measurement expressed by

the standard deviation sInstrel over the hardness.
1,4
Values for use of 2.objective

1,2 Values for use uf 1.objective
of diagonal measurement sinstrel, %
Relative reproducibility
1,0
0,8
0,6
0,4
0 100 200 300 400 500

Diagonal length d, m
Fig. 24: Relative reproducibility of the diagonal measurement expressed by the

standard deviation sInstrel over the diagonal length.
In Fig. 24 two groups of data attract the attention. These two groups of data
correspond to two different objectives used for the diagonal measurement, one with
higher magnification for smaller indents with diagonals d < 100 m and the other with
smaller magnification for indents with diagonals 250 m < d < 500 m. For both
groups the relative reproducibility sinstrel decreases with increasing diagonal length d.
But for the smaller diagonals the reproducibility increases much more sharply with
31
decreasing diagonal length than for the larger diagonals. For the smaller diagonals
(d < 100 m) the relative reproducibility is in a range sinstrel = 0,5 % to 1,3 % and for
the larger diagonals in a range sinstrel = 0,4 % to 0,7 %.
The part of the uncertainty of the diagonal measurement in the measured hardness
deviations for the two extreme scales 840 HV0,2 and 240 HV30 is depicted in Fig. 25
and 26. At this the uncertainty of diagonal measurement was converted by
d
dHV = 2 (19)
d
in hardness units. The measured hardness deviations from the arithmetic mean value
of all participants are an approximate expression of the hardness uncertainty.
60
Results for 840 HV 0,2
Measured hardness deviation
40 Hardness deviation due to
Hardness deviation H, HV
length measuring deviation
20
-20
-40
M
UM
RO
GC
B
B
L
IJ
ST
I
CM
NP
PT
PT
IS
NM
NI
NI
IM
ET
G
KR
M
IN
Fig. 25: Comparison of the measured hardness deviation from the mean value of
all participants with the hardness deviation due to length measuring deviation for the
hardness scale 840 HV0,2
18 Results for 240 HV 30

16 Measured hardness deviation
14 Hardness deviation due to
12 length measuring deviation
10
Hardness deviation H, HV
8
6
4
2
0
-2
2
10
-4
-6
-8
-10
-12
-14
M
UM
RO
GC
B
L
IJ
ST
I
CM
NP
PT
IS
NM
NI
NI
IM
ET
G
KR
M
IN
32
Fig. 26: Comparison of the measured hardness deviation from the mean value of
all participants with the hardness deviation due to length measuring deviation for the
hardness scale 240 HV30
Fig. 25 and 26 show that the hardness deviation clearly depends on the hardness
deviation due to the length measuring deviation, but this dependence becomes still
more significant for small indent sizes, as in the case of the hardness scale
840 HV0,2.
8.5 Influence of the numerical aperture of the objective

Depending on the used equipment the participants have used optical length
measuring systems with different magnifications and numerical apertures. Tab. 22
delivers an overview of the characteristics of the used optical length measuring
systems.
Table 22:
Overview of the numerical aperture of the used optical length measuring systems
Scale HV0,2 HV1 HV30
Block 240 540 840 240 540 840 240 540 840
Institute
IMGC 0,5 0,5 0,5 0,4 0,4 0,5 0,2 0,2 0,2
INMETRO - - - - - - 0,32 0,32 0,32
NIST 0,9 0,9 0,9 0,8 0,8 0,8 - - -
NMIJ 0,65 0,65 0,65 0,45 0,45 0,45 0,1 0,25 0,25
NIM 0,8 0,8 0,8 0,8 0,8 0,8 0,5 0,5 0,5
KRISS 0,65 0,65 0,65 0,65 0,4 0,4 0,3 0,25 0,25
CMI 0,5 0,5 0,5 0,5 0,5 0,5 0,35 0,35 0,35
GUM 0,32 0,32 0,32 0,32 0,32 0,32 0,3 0,3 0,3
NPL 0,9 0,9 0,9 0,8 0,9 0,9 0,9 0,9 0,9
PTB 0,75 0,75 0,75 0,75 0,75 0,75 0,2 0,2 0,2
NAmax 0,58 0,58 0,58 0,48 0,58 0,58 0,8 0,7 0,7
In the last line of Table 19 one can see that the largest difference of numerical
aperture NAmax between the participants for the different indents was between 0,58
and 0,8. Therefore one can assume that the numerical aperture of the used
objectives can influence the results of the length measurement.
In order to investigate this influence, partner IMGC made the following investigation.
The reference indents were measured with a microscope to which belongs a set of
five objectives. These objectives have the following magnifications and numerical
apertures:
Magnification 32x 20x 10x 5x 2,5x
Numerical aperture 0,5 0,4 0,2 0,09 0,07
The result of the measurement of the reference indents with the 5 different objectives
is shown in Table 23.
33
Table 23:
Result of the measurement of the reference indents with the 5 different objectives
Block Objective/N.A.
(indentations 8-8) 32x 20x 10x 5x 2.5x Ref.
0,5 0,4 0,2 0,09 0,07 0,5
240HV0,2 38,69 38,50 38,39 38,69
540HV0,2 26,15 25,94 26,15
840HV0,2 20,50 20,79 20,50
240HV1 87,77 87,47 87,29 87,77
540HV1 58,32 59,01 58,59 58,32
840HV1 46,60 46,64 46,25 46,60
240HV30 482,98 483,46 488,55 486,98
540HV30 325,04 324,96 325,35 325,46
840HV30 258,63 258,31 259,40 259,71 260,18 258,63
a
On the basis of the data in Table 19, Fig. 27 shows the relative length measuring
deviations in dependence on the diagonal length and the numerical aperture.
2,0
Relative length measuring deviation lrel, %
N.A. = 0,4
1,5 N.A. = 0,2
N.A. = 0,09
1,0
Linear fit for N.A. = 0,4
Linear fit for N.A. = 0,2
0,5
0,0
-0,5
-1,0
0 100 200 300 400 500
Mean diagonal length l, m
Fig. 27: Relative length measuring deviations in dependence on the diagonal

length and the numerical aperture (for N.A. = 0,5 the length measurement deviations
are set = 0)
Fig. 27 clarifies that the effect of numerical aperture does not have a well defined
slope and sign. Changing the objectives (and the numerical aperture), the
measurements change very much (about 1 %), but in a random way.
From this analysis it is not possible to correct for the effect of the numerical aperture.
One can only conclude that variations of the diagonal length of 1 % can be justified,
if different objectives and numerical apertures are used.
34
9 Uncertainty budgets
9.1 Results of calibrations
As basis for the calculation of the measurement uncertainty all participants carried
out a calibration of the used Vickers standard machines. The results of the
calibrations are summarised in Table 24.
Table 24:
Results of the calibrations of Vickers standard machines used by the participants
uF ud, m u, r, m c, m
HV0,2 HV1 HV30 d<40 m d>40 m
IMGC 0,02 mN 0,098 mN 0,003 N 0,01 m 0,01 m 0,008
INMETRO- - 0,15 N - 1,4 m 1 - 2 m
NIST 0,9 mN 4 mN (0,09...0,12)m(0,15...0,18)m0,075
NMIJ 0,98 mN 17 mN 0,076 N (0,22..0,28)m(0,36...0,56)m0,1 0,3 m 0,4 m
NIM 2,7 mN 13,7 mN 0,29 N 0,4 m 0,5 m 0,02 (0,5...0,6) m
KRISS 1,2 mN 5,9 mN 0,12 N 0,16 m 0,4 m 0,02 -
CMI 1 mN 1 mN 0,01 N 0,02 m 0,24 m (0,11...0,55) (0,7...0,8) m (0,8...0,9) m
GUM 1,9 mN 9,8 mN 0,29 N 0,2 m 0,5 % d <0,1 0,3 m (0,2 ... 0,3) m
NPL 0,246 N 0,14 % d 0,1 0,37 m
PTB 19,6 mN 9,8 mN 0,15 N 0,3 m 0,4 m 0,1 0,3 m 0,4 m
The data in Table 24 for force F, diagonal length d and indenter plane angle
indicate the calibration uncertainties, whereas for tip radius r and length of line of
junction c the deviation are given.
From this table one gets an overview that the calibration facilities have a rather wide
range of achievable uncertainties. Whereas the facilities for force and length are well
developed, for the characterisation of the indenter geometry in several cases no
calibration facilities are available. Here in future corresponding facilities should be
developed, set up and used.
9.2 Calculation of measurement uncertainty

As basis for the determination of the measurement uncertainty the draft guideline to
the estimation of the uncertainty of the Brinell and the Vickers measuring method
was recommended. [2] The uncertainty budgets of the participants based on this
guideline are contained in Appendix A3.
The uncertainty budgets of the participants based on the unified procedure as
presented in ch. 8.2 appear in Appendix A4.
9.3 Degree of equivalence
The degree of equivalence (DoE) of each participant with respect to the reference
value is given by DoE(dir, Uir) defined as:
d ir = d i d ref
(30)
U ir = 2 ( u i2 u ref
2
)
Here the corresponding uncertainties ui and uref cannot simply be geometrically
added, because the values di and dref are correlated.[3] The DoE is calculated for
measurement results with En< 1. Because the original method in [3] is constructed for
the case of the weighted mean value where uref is smaller than the smallest value of
ui, for the case of the arithmetic mean, which we have applied, for the radicand was
taken the absolute value in order to avoid imaginary values.
35
Then the degree of equivalence between institute i and j is calculated as follows. For
i = 1 ... N and j = 1 ... N with j not equal i, the degree of equivalence between institute
i and institute j is formed as the pair of values (dij, U(dij)) using
di , j = xi x j
(31)
U (di , j ) = 2u (d ij )
where u(di,j) is given by
u (d i , j ) = (u 2 ( xi ) + u 2 ( x j ) (32)
Discrepant measurements are identified if
di > 2u (d i ) (33)
then xi is classified as discrepant at the 5% level of significance.
The calculation of DoE according to this procedure is given in the following tables.
Table 25:
Degree of equivalence related to the reference values
1) Deviations dij from reference values

IMGC 8,03 1,30 -1,32 13,83 10,49 -0,06 23,39 1,50
INMETRO 9,43
NIST -1,40 -5,83 -13,81
NMIJ -4,69 -1,40 -4,99 -9,90 -8,63 -5,87 -7,76
NIM -3,91 -1,93 -2,75 2,30 -5,79 -6,63 -6,85 -2,61 -3,44
KRISS 1,31 -0,98 2,81 4,46 -0,46 2,90 -0,56 8,34 12,51
CMI -1,07 0,92 -0,42 -11,50 1,03 -4,78 -4,59
GUM -4,27 -1,25 3,43 -4,57 1,20 2,39 -10,81 0,36 -1,32
NPL -0,60 -0,65
PTB 6,00 3,34 -1,15 6,30 3,41 -2,62 17,26 4,38 -1,49
2) Values of uncertainty Uir

IMGC 23,27 8,87 1,45 59,45 24,76 9,09 111,38 14,50
INMETRO 18,48
NIST 19,35 22,09 27,77
NMIJ 21,98 11,04 92,34 42,27 176,86 79,15 4,69
NIM 25,62 11,26 2,16 37,98 18,55 9,51 83,53 26,95 7,46
KRISS 24,56 10,40 6,10 39,35 12,57 10,98 67,80 8,11 6,21
CMI 29,83 12,53 3,14 17,57 32,10 2,92 41,55
GUM 22,84 14,00 9,19 38,14 13,23 21,60 51,06 18,64 25,50
NPL 3,67 2,68
PTB 23,49 14,41 2,71 32,64 18,24 8,09 52,38 14,02 13,93
36
Table 26:
Hardness scale 240 HV0,2
1) Differences dij between institutes
IMGC INME- NIST NMIJ NIM KRISS CMI GUM NPL PTB
TRO
IMGC 0 0 9,42 12,71 11,94 6,72 9,09 12,16 0 2,03
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -9,42 0 0,00 3,29 2,52 -2,70 -0,33 2,74 0 -7,39
NMIJ -12,71 0 -3,29 0 -0,77 -5,99 -3,62 -0,55 0 -10,68
NIM -11,94 0 -2,52 0,77 0 -5,22 -2,85 0,22 0 -9,91
KRISS -6,72 0 2,70 5,99 5,22 0 2,37 5,44 0 -4,69
CMI -9,09 0 0,33 3,62 2,85 -2,37 0 3,07 0 -7,06
GUM -12,16 0 -2,74 0,55 -0,22 -5,44 -3,07 0 0 -10,13
NPL 0 0 0 0 0 0 0 0 0 0
PTB -2,03 0 7,39 10,68 9,91 4,69 7,06 10,13 0 0
2) Uncertainty between institutes U(dij)
IMGC INMETRO NIST NMIJ NIM KRISS CMI GUM NPL PTB
IMGC 0 0 32,91 34,52 36,94 36,21 39,98 35,07 0 35,50
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 32,01 34,61 37,83 37,83 32,61 0 33,06
NMIJ 0 36,14 35,40 39,24 34,23 0 34,66
NIM 0 37,77 41,39 36,67 0 37,08
KRISS 0 40,74 35,94 0 36,35
CMI 0 39,73 0 40,11
GUM 0 0 35,22
NPL 0 0
PTB 0
37
Table 27:
Hardness scale 240 HV1
IMGC 0 0 0 2,70 3,23 2,28 0,38 2,55 0 -2,04
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -2,70 0 0 0 0,53 -0,42 -2,32 -0,15 0 -4,74
NIM -3,23 0 0 -0,53 0 -0,95 -2,85 -0,68 0 -5,27
KRISS -2,28 0 0 0,42 0,95 0 -1,90 0,27 0 -4,32
CMI -0,38 0 0 2,32 2,85 1,90 0 2,17 0 -2,42
GUM -2,55 0 0 0,15 0,68 -0,27 -2,17 0 0 -4,59
NPL 0 0 0 0 0 0 0 0 0 0
PTB 2,04 0 0 4,74 5,27 4,32 2,42 4,59 0 0
IMGC 0 0 9,80 15,34 15,49 14,88 16,44 17,58 0 17,92
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 16,83 16,27 17,71 18,77 0 19,09
NIM 0 16,41 17,84 18,90 0 19,21
KRISS 0 17,31 18,40 0 18,72
CMI 0 19,68 0 19,98
GUM 0 0 20,93
NPL 0 0
PTB 0
Table 28:
IMGC 0 0 0 0 1,43 -4,13 -0,90 -4,75 -0,72 -0,17
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0 0 0 0
NIM -1,43 0 0 0 0 -5,56 -2,33 -6,18 -2,15 -1,60
KRISS 4,13 0 0 0 5,56 0 3,23 -0,62 3,41 3,96
CMI 0,90 0 0 0 2,33 -3,23 0 -3,85 0,18 0,73
GUM 4,75 0 0 0 6,18 0,62 3,85 0 4,03 4,58
NPL 0,72 0 0 0 2,15 -3,41 -0,18 -4,03 0 0,55
PTB 0,17 0 0 0 1,60 -3,96 -0,73 -4,58 -0,55 0
38
IMGC 0 0 0 0 5,34 8,39 4,83 10,85 4,44 5,08
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0
NIM 0 7,97 4,07 10,53 3,60 4,37
KRISS 0 7,64 12,36 7,40 7,80
CMI 0 10,28 2,78 3,72
GUM 0 10,10 10,40
NPL 0 3,20
PTB 0
Table 29:
IMGC 0 0 19,66 18,82 11,53 9,37 25,33 18,40 0 7,53
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -19,66 0 0,00 -0,84 -8,13 -10,29 5,67 -1,26 0 -12,13
NMIJ -18,82 0 0,84 0 -7,29 -9,45 6,51 -0,42 0 -11,29
NIM -11,53 0 8,13 7,29 0 -2,16 13,80 6,87 0 -4,00
KRISS -9,37 0 10,29 9,45 2,16 0 15,96 9,03 0 -1,84
CMI -25,33 0 -5,67 -6,51 -13,80 -15,96 0 -6,93 0 -17,80
GUM -18,40 0 1,26 0,42 -6,87 -9,03 6,93 0 0 -10,87
NPL 0 0 0 0 0 0 0 0 0 0
PTB -7,53 0 12,13 11,29 4,00 1,84 17,80 10,87 0 0
IMGC 0 0 68,13 112,61 74,81 75,51 66,80 74,89 0 72,24
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 98,15 50,50 28,24 37,62 50,61 0 46,61
NMIJ 0 102,90 103,41 97,23 102,96 0 101,05
NIM 0 60,08 48,69 59,30 0 55,92
KRISS 0 49,76 60,18 0 56,86
CMI 0 48,81 0 44,64
GUM 0 0 56,03
NPL 0 0
PTB 0
39
Table 30:
IMGC 0 0 0 20,39 16,28 10,95 9,46 9,29 0 7,08
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -20,39 0 0 0 -4,11 -9,44 -10,93 -11,10 0 -13,31
NIM -16,28 0 0 4,11 0 -5,33 -6,82 -6,99 0 -9,20
KRISS -10,95 0 0 9,44 5,33 0 -1,49 -1,66 0 -3,87
CMI -9,46 0 0 10,93 6,82 1,49 0 -0,17 0 -2,38
GUM -9,29 0 0 11,10 6,99 1,66 0,17 0 0 -2,21
NPL 0 0 0 0 0 0 0 0 0 0
PTB -7,08 0 0 13,31 9,20 3,87 2,38 2,21 0 0
IMGC 0 0 0 51,94 35,42 32,69 44,06 32,95 0 35,26
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 49,28 47,36 55,81 47,54 0 49,17
NIM 0 28,28 40,89 28,58 0 31,21
KRISS 0 38,55 25,12 0 28,08
CMI 0 38,77 0 40,75
GUM 0 0 28,38
NPL 0 0
PTB 0
Table 31:
IMGC 0 -4,47 0 0 6,57 -2,96 4,72 -2,45 0,59 2,56
INMETRO 9,49 0 0 0 16,06 6,53 14,21 7,04 10,08 12,05
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0 0 0 0
NIM -6,57 -16,06 0 0 0 -9,53 -1,85 -9,02 -5,98 -4,01
KRISS 2,96 -6,53 0 0 9,53 0 7,68 0,51 3,55 5,52
CMI -4,72 -14,21 0 0 1,85 -7,68 0 -7,17 -4,13 -2,16
GUM 2,45 -7,04 0 0 9,02 -0,51 7,17 0 3,04 5,01
NPL -0,59 -10,08 0 0 5,98 -3,55 4,13 -3,04 0 1,97
PTB -2,56 -12,05 0 0 4,01 -5,52 2,16 -5,01 -1,97 0
40
IMGC 0 21,42 0 0 14,41 15,42 11,22 24,17 10,49 13,52
INMETRO 0 0 0 21,60 22,28 19,61 29,03 19,21 21,01
NIST 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0
NIM 0 15,67 11,56 24,33 10,86 13,80
KRISS 0 12,79 24,94 12,16 14,85
CMI 0 22,58 5,99 10,42
GUM 0 22,23 23,81
NPL 0 9,64
PTB 0
Table 32:
IMGC 0 0 37,20 32,02 30,24 23,95 0 34,20 0 6,13
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -37,20 0 0 -5,18 -6,96 -13,25 0 -3,00 0 -31,07
NMIJ -32,02 0 5,18 0 -1,78 -8,07 0 2,18 0 -25,89
NIM -30,24 0 6,96 1,78 0 -6,29 0 3,96 0 -24,11
KRISS -23,95 0 13,25 8,07 6,29 0 0 10,25 0 -17,82
CMI 0,00 0 0 0 0 0 0 0 0 0
GUM -34,20 0 3,00 -2,18 -3,96 -10,25 0 0 0 -28,07
NPL 0 0 0 0 0 0 0 0 0 0
PTB -6,13 0 31,07 25,89 24,11 17,82 0 28,07 0 0
IMGC 0 0 135,39 220,99 156,64 148,85 0 142,01 0 142,49
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 192,88 113,58 57,86 0 92,37 0 93,10
NMIJ 0 208,35 202,56 0 197,59 0 197,93
NIM 0 129,33 0 121,40 0 121,96
KRISS 0 0 111,17 0 111,78
CMI 0 0 0 0
GUM 0 0 102,49
NPL 0 0
PTB 0
41
Table 33:
IMGC 0 0 0 0 0 0 0 0 0 0
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 -3,26 -14,21 -1,28 -6,23 0 -10,25
NIM 0 0 0 3,26 0 -10,95 1,98 -2,97 0 -6,99
KRISS 0 0 0 14,21 10,95 0 12,93 7,98 0 3,96
CMI 0 0 0 1,28 -1,98 -12,93 0 -4,95 0 -8,97
GUM 0 0 0 6,23 2,97 -7,98 4,95 0 0 -4,02
NPL 0 0 0 0 0 0 0 0 0 0
PTB 0 0 0 10,25 6,99 -3,96 8,97 4,02 0 0
IMGC 0 0 0 0 0 0 0 0 0 0
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 96,20 92,70 101,26 90,44 86,00 93,40
NIM 0 55,27 68,67 51,40 43,10 56,44
KRISS 0 63,68 44,51 34,60 50,25
CMI 0 60,35 53,46 64,70
GUM 0 28,00 45,96
NPL 0 0,00
PTB 0
Table 34:
IMGC 0 0 0 9,26 4,94 -11,01 0 2,82 0 2,99
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -9,26 0 0 0 -4,32 -20,27 0 -6,44 0 -6,27
NIM -4,94 0 0 4,32 0 -15,95 0 -2,12 0 -1,95
KRISS 11,01 0 0 20,27 15,95 0 0 13,83 0 14,00
CMI 0 0 0 0 0 0 0 0 0 0
GUM -2,82 0 0 6,44 2,12 -13,83 0 0 0 0,17
NPL 0 0 0 0 0 0 0 0 0 0
PTB -2,99 0 0 6,27 1,95 -14,00 0 -0,17 0 0
42
IMGC 0 0 0 27,93 27,32 27,63 0 38,10 0 24,65
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 30,57 30,85 0 40,50 0 28,21
NIM 0 30,29 0 40,08 0 27,61
KRISS 0 0 40,29 0 27,92
CMI 0 0 0 0
GUM 0 0 38,31
NPL 0 0
PTB 0
The consistency check delivered the result that the tables 26 to 34 do not contain
discrepant values for DoE. As expected this consistency check concurs with the En
values.
10 Discussions, conclusions and remarks
The CCM Vickers key comparison can be considered as a successful metrological

exercise. Representatively it delivered for all three ranges of the test forces (Micro-
Vickers, small load Vickers and Macro-Vickers scales) valuable metrological data. At
present Vickers hardness reference blocks with high time-dependent stability and
high local homogeneity, including high surface quality are available.
The uncertainties of the reference values are obviously the smallest uncertainties
reached in the field of Vickers measurements worldwide so far. These uncertainties
can be interpreted as the present limits of Vickers measurements in the investigated
range of hardness scales. This is one important outcome of this Vickers key
comparison.
In order to overcome these metrological limits in the future, it is recommended to
concentrate metrological investigations on the following topics:
1) It was found that the calibration methods for the diagonal measurements,
especially for diagonal lengths d > 100 m should be improved.
2) The calibration methods for the parameters of the indenter geometry, like tip
radius and length of the line of junction should be further developed.
3) The inputs from the participants to the used draft guideline for the estimation of the
measurement uncertainty should be used for its further qualification. In this context it
seems to be necessary to carry out experimental investigations on the determination
of sensitivity coefficients for material dependent influences on the Vickers hardness
measurement, especially influences of the force-time regime on the Vickers
hardness.
4) For the further reduction of the uncertainty of Micro-Vickers measurements it is
necessary to correct for the indenter deviations and to provide indenters with higher
quality.
5) With the development of automatic Vickers indent measurements on the basis of
CCD technique it is necessary to guarantee a high agreement of measurements both
with CCD systems and with optical microscopes. This requires investigations about
the relationships between both methods.
43
6) For the diagonal measurements with optical microscopes the properties of the
used optical system should be further investigated.
11 References
[1] T. J. Quinn, Guidelines for key comparisons carried out by Consultative

Committees, BIPM, Paris
[2] EA Working group Hardness; Draft: Guideline to the estimation of the
uncertainty of the Brinell and the Vickers measuring method, July 2002
[3] W. Bich, M.G. Cox, W.T. Estler, L. Nielsen, W. Woeger: Proposed guidelines for
the evaluation of key comparison data, BIPM (2002)
44
Abstract
In the framework of the Working Group on Hardness (WGH) of the Consultative
Committee for Mechanical Measurements (CCM) in the year 2003 the key
comparison Vickers was finished. In the comparison the hardness laboratories of the
following national metrology institutes participated: IMGC (Italy), NIST (USA),
INMETRO (Brazil), NIM (P.R. China), KRISS (Republic of Korea), NMIJ (Japan), CMI
(Czech Republic), GUM (Poland), NPL (U.K.), and PTB (Germany) served as the
pilot laboratory. The comparison of the Vickers primary hardness standard machines
was carried out with three sets of hardness reference blocks of the Vickers scales
HV0,2, HV1 and HV 30 each with the hardness levels 240 HV, 540 HV and 840 HV.
The Vickers key comparison for all used test forces ranges (Micro Vickers, Small
force Vickers, Macro Vickers) delivered the following results.
Table A1:
240 HV
HV0,2 HV1 HV30

Institute Mean value Std.dev. Mean value Std. dev. Mean value . Std. dev.
PTB 248,19 2,67 244,06 1,79
IMGC 254,30 4,34 241,11 3,18 237,83 0,88
INMETRO 244,83 0,77
NIST 244,88 3,31 236,13 1,81
NMIJ 241,59 3,92 238,41 1,71 236,11 1,21
NIM 242,36 2,77 237,88 0,86 236,40 0,44
KRISS 247,58 3,48 238,83 1,88 241,96 0,46
CMI 245,21 3,00 240,73 0,66 238,73 0,39
GUM 242,14 2,56 238,56 2,28 242,58 0,77
NPL 253,65 2,92 240,33 3,55 238,55 0,53
PTB 252,27 2,99 243,15 2,46 238,00 0,56
Mean value 247,20 239,92 239,44

Std.dev. 4,83 2,44 2,99
Table A2:
540 HV
HV0,2 HV1 HV30

Institute Mean val.x8 Std.dev.s8 Mean val.x8 Std. dev.s8 Mean val.x8 Std. dev.s8
PTB 528,42 5,81 539,03 5,97
IMGC 548,16 6,12 544,84 5,71 525,45 1,35
INMETRO 534,94 2,08
NIST 528,50 3,34 522,50 1,77
NMIJ 529,34 6,24 524,45 5,71 516,68 1,90
NIM 536,63 1,87 528,56 2,10 518,88 0,79
KRISS 538,79 4,61 533,89 2,94 528,41 1,57
CMI 522,83 5,50 535,38 1,30 520,73 0,57
GUM 529,76 6,11 535,55 2,72 527,90 1,30
NPL 547,36 13,07 536,63 4,34 524,86 1,06
PTB 540,63 5,22 537,76 1,59 522,89 2,50
M.val.xInst 535,04 533,86 524,53

Std.dev.sInst 45
8,61 6,84 5,55
Table A3:
840 HV

Results for 840 HV hardness reference blocks
HV0,2 HV1 HV30
Institute Mean val.x8 Std.dev. s8 Mean val.x8 Std. dev. s8 Mean val.x8 Std. dev. s8
PTB 839,78 9,38 831,49 6,14
IMGC 864,20 44,79 875,40 10,05 820,85 2,15
INMETRO 841,55 1,43
NIST 827,00 7,09 815,38 6,35
NMIJ 832,18 13,40 825,54 7,39 811,59 5,93
NIM 833,96 5,00 828,80 3,24 815,91 1,71
KRISS 840,25 3,77 839,75 5,77 831,86 1,68
CMI 813,37 4,02 826,82 4,45 810,64 1,08
GUM 830,00 5,77 831,77 3,26 818,03 2,20
NPL 897,57 13,60 839,32 7,54 837,17 1,55
PTB 858,07 10,22 835,79 6,62 817,86 3,39
M.val.xInst 843,64 835,01 822,83

Std.dev. sInst 23,97 15,91 11,25
Table A4:
by the unified procedure)
Hardness Reference value xref, Uncertainty of reference value
scale HV Uref
Abs. unc., HV Rel. unc., %
240 HV0,2 246,28 4,57 1,86
240 HV1 239,81 2,08 0,87
240 HV30 239,15 1,97 0,82
540 HV0,2 534,33 8,80 1,65
540 HV1 534,35 6,10 1,14
540 HV30 525,51 5,03 0,96
840 HV0,2 840,81 25,38 3,02
840 HV1 831,41 16,82 2,02
840 HV30 819,35 11,25 1,37
46
Table A5:
Results of the relative measurement uncertainty (in %) according to the estimation of
the participant (part) and to the unified procedure (unif)
Institute 240HV0,2 240HV1 240HV30 540HV0,2 540HV1 540HV30 840HV0,2 840HV1 840HV30
part unif part unif part unif part unif part unif part unif part unif part unif part unif
IMGC 3,9 4,1 1,7 1,9 0,9 0,7 5,7 5,8 2,5 2,8 1,0 0,8 7,1 7,0 3,2 3,4 1,0 0,9
INMETRO 1,8 4,1 1,8 3,7 1,8 4,6
NIST 2,0 2,7 1,1 1,2 2,7 4,0 1,4 1,9 3,5 5,1 1,6 2,2
NMIJ 5,8 3,3 2,9 1,8 0,9 0,6 8,9 4,9 4,2 2,6 1,2 1,0 11,1 6,1 5,2 3,3 1,4 1,2
NIM 3,1 4,5 1,4 2,1 0,7 0,6 3,9 6,9 2,1 3,2 1,0 1,0 5,9 8,6 2,6 3,9 1,3 1,2
KRISS 2,9 2,4 1,4 1,1 1,5 0,6 4,0 3,8 1,6 1,4 1,1 0,7 5,0 4,1 2,1 2,3 1,3 0,7
CMI 2,4 1,7 3,2 0,7 0,5 0,6 2,3 2,5 3,2 1,0 0,5 0,7 2,2 3,1 3,2 1,3 0,5 0,8
GUM 3,3 2,9 1,7 2,4 2,1 2,3 3,9 4,3 1,7 2,8 2,0 2,2 4,3 5,4 1,7 2,6 2,1 2,2
NPL 0,3 0,6 0,3 0,7 0,3 0,8
PTB 3,1 3,3 2,0 2,3 0,6 0,6 3,4 3,0 2,0 2,2 0,9 1,0 4,3 3,7 2,2 2,2 1,1 1,2
Mean value 3,3 3,1 1,9 1,7 1,0 1,2 4,4 4,4 2,3 2,2 1,1 1,3 5,4 5,4 2,7 2,7 1,2 1,5
Variation 3,8 2,8 2,1 1,7 1,8 3,5 6,6 4,4 2,8 2,2 1,7 3 8,9 5,5 3,6 2,6 1,8 3,9
Table A6:
measurement uncertainty (based on the arithmetic mean)
En values
240 HV 540 HV 840 HV
IMGC 0,55 0,32 0,62 0,31 0,63 0,17 0,26 1,06 0,18
INMETRO 0,66 0,84 0,76
NIST 0,19 0,67 0,20 0,63 0,24 0,62
NMIJ 0,44 0,17 1,12 0,16 0,44 1,10 0,15 0,24 0,79
NIM 0,33 0,25 1,02 0,02 0,22 0,77 0,11 0,16 0,48
KRISS 0,04 0,11 0,78 0,07 0,03 0,68 0,05 0,11 0,70
CMI 0,12 0,19 0,22 0,26 0,09 0,51 0,31 0,19 0,85
GUM 0,39 0,12 0,48 0,15 0,11 0,24 0,17 0,11 0,22
NPL 0,31 0,05 1,12
PTB 0,39 0,48 0,45 0,16 0,24 0,20 0,23 0,01 0,31
Mean 0,31 0,29 0,63 0,17 0,30 0,51 0,19 0,31 0,60
The degree of equivalence between the participating institutes for the different
hardness scales follows from the tables A7 to A16.
Table A7:
Degree of equivalence related to the reference values
1) Deviations dij from reference values

IMGC 8,03 1,30 -1,32 13,83 10,49 -0,06 23,39 1,50
INMETRO 9,43
NIST -1,40 -5,83 -13,81
NMIJ -4,69 -1,40 -4,99 -9,90 -8,63 -5,87 -7,76
NIM -3,91 -1,93 -2,75 2,30 -5,79 -6,63 -6,85 -2,61 -3,44
KRISS 1,31 -0,98 2,81 4,46 -0,46 2,90 -0,56 8,34 12,51
CMI -1,07 0,92 -0,42 -11,50 1,03 -4,78 -4,59
GUM -4,27 -1,25 3,43 -4,57 1,20 2,39 -10,81 0,36 -1,32
NPL -0,60 -0,65
PTB 6,00 3,34 -1,15 6,30 3,41 -2,62 17,26 4,38 -1,49
47
2) Values of uncertainty Uir

IMGC 23,27 8,87 1,45 59,45 24,76 9,09 111,38 14,50
INMETRO 18,48
NIST 19,35 22,09 27,77
NMIJ 21,98 11,04 92,34 42,27 176,86 79,15 4,69
NIM 25,62 11,26 2,16 37,98 18,55 9,51 83,53 26,95 7,46
KRISS 24,56 10,40 6,10 39,35 12,57 10,98 67,80 8,11 6,21
CMI 29,83 12,53 3,14 17,57 32,10 2,92 41,55
GUM 22,84 14,00 9,19 38,14 13,23 21,60 51,06 18,64 25,50
NPL 3,67 2,68
PTB 23,49 14,41 2,71 32,64 18,24 8,09 52,38 14,02 13,93
Table A8:
IMGC INME- NIST NMIJ NIM KRISS CMI GUM NPL PTB
TRO
IMGC 0 0 9,42 12,71 11,94 6,72 9,09 12,16 0 2,03
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -9,42 0 0,00 3,29 2,52 -2,70 -0,33 2,74 0 -7,39
NMIJ -12,71 0 -3,29 0 -0,77 -5,99 -3,62 -0,55 0 -10,68
NIM -11,94 0 -2,52 0,77 0 -5,22 -2,85 0,22 0 -9,91
KRISS -6,72 0 2,70 5,99 5,22 0 2,37 5,44 0 -4,69
CMI -9,09 0 0,33 3,62 2,85 -2,37 0 3,07 0 -7,06
GUM -12,16 0 -2,74 0,55 -0,22 -5,44 -3,07 0 0 -10,13
NPL 0 0 0 0 0 0 0 0 0 0
PTB -2,03 0 7,39 10,68 9,91 4,69 7,06 10,13 0 0
IMGC 0 0 32,91 34,52 36,94 36,21 39,98 35,07 0 35,50
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 32,01 34,61 37,83 37,83 32,61 0 33,06
NMIJ 0 36,14 35,40 39,24 34,23 0 34,66
NIM 0 37,77 41,39 36,67 0 37,08
KRISS 0 40,74 35,94 0 36,35
CMI 0 39,73 0 40,11
GUM 0 0 35,22
NPL 0 0
PTB 0
48
Table A9:
IMGC 0 0 0 2,70 3,23 2,28 0,38 2,55 0 -2,04
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -2,70 0 0 0 0,53 -0,42 -2,32 -0,15 0 -4,74
NIM -3,23 0 0 -0,53 0 -0,95 -2,85 -0,68 0 -5,27
KRISS -2,28 0 0 0,42 0,95 0 -1,90 0,27 0 -4,32
CMI -0,38 0 0 2,32 2,85 1,90 0 2,17 0 -2,42
GUM -2,55 0 0 0,15 0,68 -0,27 -2,17 0 0 -4,59
NPL 0 0 0 0 0 0 0 0 0 0
PTB 2,04 0 0 4,74 5,27 4,32 2,42 4,59 0 0
IMGC 0 0 9,80 15,34 15,49 14,88 16,44 17,58 0 17,92
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 16,83 16,27 17,71 18,77 0 19,09
NIM 0 16,41 17,84 18,90 0 19,21
KRISS 0 17,31 18,40 0 18,72
CMI 0 19,68 0 19,98
GUM 0 0 20,93
NPL 0 0
PTB 0
Table A10:
IMGC 0 0 0 0 1,43 -4,13 -0,90 -4,75 -0,72 -0,17
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0 0 0 0
NIM -1,43 0 0 0 0 -5,56 -2,33 -6,18 -2,15 -1,60
KRISS 4,13 0 0 0 5,56 0 3,23 -0,62 3,41 3,96
CMI 0,90 0 0 0 2,33 -3,23 0 -3,85 0,18 0,73
GUM 4,75 0 0 0 6,18 0,62 3,85 0 4,03 4,58
NPL 0,72 0 0 0 2,15 -3,41 -0,18 -4,03 0 0,55
PTB 0,17 0 0 0 1,60 -3,96 -0,73 -4,58 -0,55 0
49
IMGC 0 0 0 0 5,34 8,39 4,83 10,85 4,44 5,08
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0
NIM 0 7,97 4,07 10,53 3,60 4,37
KRISS 0 7,64 12,36 7,40 7,80
CMI 0 10,28 2,78 3,72
GUM 0 10,10 10,40
NPL 0 3,20
PTB 0
Table A11:
IMGC 0 0 19,66 18,82 11,53 9,37 25,33 18,40 0 7,53
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -19,66 0 0,00 -0,84 -8,13 -10,29 5,67 -1,26 0 -12,13
NMIJ -18,82 0 0,84 0 -7,29 -9,45 6,51 -0,42 0 -11,29
NIM -11,53 0 8,13 7,29 0 -2,16 13,80 6,87 0 -4,00
KRISS -9,37 0 10,29 9,45 2,16 0 15,96 9,03 0 -1,84
CMI -25,33 0 -5,67 -6,51 -13,80 -15,96 0 -6,93 0 -17,80
GUM -18,40 0 1,26 0,42 -6,87 -9,03 6,93 0 0 -10,87
NPL 0 0 0 0 0 0 0 0 0 0
PTB -7,53 0 12,13 11,29 4,00 1,84 17,80 10,87 0 0
IMGC 0 0 68,13 112,61 74,81 75,51 66,80 74,89 0 72,24
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 98,15 50,50 28,24 37,62 50,61 0 46,61
NMIJ 0 102,90 103,41 97,23 102,96 0 101,05
NIM 0 60,08 48,69 59,30 0 55,92
KRISS 0 49,76 60,18 0 56,86
CMI 0 48,81 0 44,64
GUM 0 0 56,03
NPL 0 0
PTB 0
50
Table A12:
IMGC 0 0 0 20,39 16,28 10,95 9,46 9,29 0 7,08
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -20,39 0 0 0 -4,11 -9,44 -10,93 -11,10 0 -13,31
NIM -16,28 0 0 4,11 0 -5,33 -6,82 -6,99 0 -9,20
KRISS -10,95 0 0 9,44 5,33 0 -1,49 -1,66 0 -3,87
CMI -9,46 0 0 10,93 6,82 1,49 0 -0,17 0 -2,38
GUM -9,29 0 0 11,10 6,99 1,66 0,17 0 0 -2,21
NPL 0 0 0 0 0 0 0 0 0 0
PTB -7,08 0 0 13,31 9,20 3,87 2,38 2,21 0 0
IMGC 0 0 0 51,94 35,42 32,69 44,06 32,95 0 35,26
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 49,28 47,36 55,81 47,54 0 49,17
NIM 0 28,28 40,89 28,58 0 31,21
KRISS 0 38,55 25,12 0 28,08
CMI 0 38,77 0 40,75
GUM 0 0 28,38
NPL 0 0
PTB 0
Table A13:
IMGC 0 -4,47 0 0 6,57 -2,96 4,72 -2,45 0,59 2,56
INMETRO 9,49 0 0 0 16,06 6,53 14,21 7,04 10,08 12,05
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0 0 0 0
NIM -6,57 -16,06 0 0 0 -9,53 -1,85 -9,02 -5,98 -4,01
KRISS 2,96 -6,53 0 0 9,53 0 7,68 0,51 3,55 5,52
CMI -4,72 -14,21 0 0 1,85 -7,68 0 -7,17 -4,13 -2,16
GUM 2,45 -7,04 0 0 9,02 -0,51 7,17 0 3,04 5,01
NPL -0,59 -10,08 0 0 5,98 -3,55 4,13 -3,04 0 1,97
PTB -2,56 -12,05 0 0 4,01 -5,52 2,16 -5,01 -1,97 0
51
IMGC 0 21,42 0 0 14,41 15,42 11,22 24,17 10,49 13,52
INMETRO 0 0 0 21,60 22,28 19,61 29,03 19,21 21,01
NIST 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 0 0 0
NIM 0 15,67 11,56 24,33 10,86 13,80
KRISS 0 12,79 24,94 12,16 14,85
CMI 0 22,58 5,99 10,42
GUM 0 22,23 23,81
NPL 0 9,64
PTB 0
Table A14:
IMGC 0 0 37,20 32,02 30,24 23,95 0 34,20 0 6,13
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST -37,20 0 0 -5,18 -6,96 -13,25 0 -3,00 0 -31,07
NMIJ -32,02 0 5,18 0 -1,78 -8,07 0 2,18 0 -25,89
NIM -30,24 0 6,96 1,78 0 -6,29 0 3,96 0 -24,11
KRISS -23,95 0 13,25 8,07 6,29 0 0 10,25 0 -17,82
CMI 0 0 0 0 0 0 0 0 0 0
GUM -34,20 0 3,00 -2,18 -3,96 -10,25 0 0 0 -28,07
NPL 0 0 0 0 0 0 0 0 0 0
PTB -6,13 0 31,07 25,89 24,11 17,82 0 28,07 0 0
IMGC 0 0 135,39 220,99 156,64 148,85 0 142,01 0 142,49
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 192,88 113,58 57,86 0 92,37 0 93,10
NMIJ 0 208,35 202,56 0 197,59 0 197,93
NIM 0 129,33 0 121,40 0 121,96
KRISS 0 0 111,17 0 111,78
CMI 0 0 0 0
GUM 0 0 102,49
NPL 0 0
PTB 0
52
Table A15:
IMGC 0 0 0 0 0 0 0 0 0 0
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ 0 0 0 0 -3,26 -14,21 -1,28 -6,23 0 -10,25
NIM 0 0 0 3,26 0 -10,95 1,98 -2,97 0 -6,99
KRISS 0 0 0 14,21 10,95 0 12,93 7,98 0 3,96
CMI 0 0 0 1,28 -1,98 -12,93 0 -4,95 0 -8,97
GUM 0 0 0 6,23 2,97 -7,98 4,95 0 0 -4,02
NPL 0 0 0 0 0 0 0 0 0 0
PTB 0 0 0 10,25 6,99 -3,96 8,97 4,02 0 0
IMGC 0 0 0 0 0 0 0 0 0 0
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 96,20 92,70 101,26 90,44 86,00 93,40
NIM 0 55,27 68,67 51,40 43,10 56,44
KRISS 0 63,68 44,51 34,60 50,25
CMI 0 60,35 53,46 64,70
GUM 0 28,00 45,96
NPL 0 0,00
PTB 0
Table A16:
IMGC 0 0 0 9,26 4,94 -11,01 0 2,82 0 2,99
INMETRO 0 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0 0 0
NMIJ -9,26 0 0 0 -4,32 -20,27 0 -6,44 0 -6,27
NIM -4,94 0 0 4,32 0 -15,95 0 -2,12 0 -1,95
KRISS 11,01 0 0 20,27 15,95 0 0 13,83 0 14,00
CMI 0 0 0 0 0 0 0 0 0 0
GUM -2,82 0 0 6,44 2,12 -13,83 0 0 0 0,17
NPL 0 0 0 0 0 0 0 0 0 0
PTB -2,99 0 0 6,27 1,95 -14,00 0 -0,17 0 0
53
IMGC 0 0 0 27,93 27,32 27,63 0 38,10 0 24,65
INMETRO 0 0 0 0 0 0 0 0 0
NIST 0 0 0 0 0 0 0 0
NMIJ 0 30,57 30,85 0 40,50 0 28,21
NIM 0 30,29 0 40,08 0 27,61
KRISS 0 0 40,29 0 27,92
CMI 0 0 0 0
GUM 0 0 38,31
NPL 0 0
PTB 0
The uncertainties of the reference values can be considered as the present accuracy
limits of Vickers measurements in the investigated ranges of hardness scales.
In order to beat these metrological limits in the future, it is recommended to
concentrate metrological investigations mainly on the following topics:
1) Improvement of the calibration methods for the diagonal measurements for
diagonal lengths d > 100 m
2) Further development of the calibration methods for the parameters of the
indenter geometry, like tip radius and length of the line of junction
3) Experimental investigations for the determination of sensitivity coefficients
for material dependent influences on Vickers measurements
4) Investigation of the relationships between results of Vickers measurements
received with CCD systems and conventional optical microscopes
5) Further qualification of the guideline for the estimation of the uncertainty of
Vickers measurements
The proposed investigations should be realised in a co-ordinated way in the
framework of the WGH.
54

CCM Vickers Key Comparison Final Report: Physikalisch-Technische Bundesanstalt Braunschweig

Uploaded by

Copyright:

Available Formats

CCM Vickers Key Comparison Final Report: Physikalisch-Technische Bundesanstalt Braunschweig

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CCM Vickers Key Comparison Final Report: Physikalisch-Technische Bundesanstalt Braunschweig

Uploaded by

Copyright:

Available Formats

Physikalisch-Technische Bundesanstalt Braunschweig

CCM Vickers key comparison

Braunschweig, March 2005 / K. Herrmann

6 Stability of the standards 7

8.1 Reference values on the basis of the uncertainty determined by the

10 Discussions, conclusions and remarks 42

The metrological equivalence of national measurement standards is determined by a

Participant Institute Adress

2.2 Time schedule and measurements carried out by the participants

Table 2: Time schedule

Laboratory Original schedule Confirmation of Results received

Laboratory HV0,2 HV1 HV30 Ref. Ref. Ref.

6 Stability of the standards

Hardness reference Hardness change, HV Relative hardness change,

HV0,2 HV1 HV30

Mean value 247,20 239,92 239,44

HV0,2 HV1 HV30

M.val.xInst 535,04 533,86 524,53

M.val.xInst 843,64 835,01 822,83

8.1.1 Reference values based on the arithmetic mean

Hardness scale Ref. value xref Uncertainty of reference value Uref

8.1.2 Reference values based on the weighted mean

U xw = (u1 S ) 2 + (u 2 S ) 2 + ... + (u n S ) 2 (5)

Hardness scale Ref. value xref Uncertainty of reference value Uref

Fig. 5: Dependency of the relative uncertainty of the reference values on the

270 240 HV0,2

580 540 HV0,2

8.2 Reference values on the basis of a unified estimation of the hardness

Influencing quantity X i Symbol Unit Value xi si ai u 2 (x i ) ci u 2 (y i ) i u i 4 (y)/ i

From the influencing quantities Xi measurement deviations xi and uncertainties in

Sum of hardness deviations:

Effective degrees of freedom, according to the Welch-Satterthwaite formula:

Hardness Reference value xref, Uncertainty of reference value

120 840 HV0,2

8.3 Influence of the inhomogeneity of hardness of the hardness reference

Further, the standard deviation due to inhomogeneity of hardness is expressed by:

Table 18 summarises the results of the inhomogeneity of the hardness reference

Standard deviation due to inhomogeneity s Inh

8.4 Influence of the length measuring deviation

The standard deviation sInst represents the reproducibility of measurements between

HV0,2 HV1 HV30

Mean value 38,71 0,13 87,89 0,12 483,11 0,38

Mean value 26,20 0,09 58,83 0,09 325,68 0,23

HV0,2 HV1 HV30

Mean value 20,99 0,11 47,16 0,18 259,28 0,20

Fig. 24 shows the relative reproducibility of the diagonal measurement expressed by

Values for use of 2.objective

0 100 200 300 400 500

Fig. 24: Relative reproducibility of the diagonal measurement expressed by the

length measuring deviation

18 Results for 240 HV 30

8.5 Influence of the numerical aperture of the objective

Fig. 27: Relative length measuring deviations in dependence on the diagonal

9.2 Calculation of measurement uncertainty

9.3 Degree of equivalence

240HV0,2 240HV1 240HV30 540HV0,2 540HV1 540HV30 840HV0,2 840HV1 840HV30

2) Values of uncertainty Uir

240HV0,2 240HV1 240HV30 540HV0,2 540HV1 540HV30 840HV0,2 840HV1 840HV30