Professional Documents
Culture Documents
Tensile Testing of Metals Proficiency Testing Program Round 13 July 2020
Tensile Testing of Metals Proficiency Testing Program Round 13 July 2020
1208
ACKNOWLEDGMENTS
PTA wishes to gratefully acknowledge the technical assistance provided
for this program by Mr S Sameem, ARL Laboratory Services Pty Ltd.
Also our thanks go to ARL Laboratory Services Pty Ltd, for the supply
and homogeneity testing of the samples.
1. FOREWORD 1
5. OUTLIER RESULTS 4
Table A: Summary Statistics for All Tests 4
Table B: Summary of Statistical Outliers 5
7. REFERENCES 13
APPENDICES
APPENDIX A
Summary of Results
0.2% Proof Stress A1.1
Lower Yield A2.1
Upper Yield A3.1
Tensile Strength A4.1
Percentage Elongation after Fracture A5.1
Percentage Reduction in Area after Fracture A6.1
Method Information and Other Reported Results A7.1
APPENDIX B
Homogeneity Testing B1.1
APPENDIX C
Instructions to Participants C1.1
Results Sheet C2.1
-1-
1. FOREWORD
(a) A total of 24 laboratories participated in the program, two of which did not
return results for inclusion in the final report. Laboratories from the following
countries received samples:
7 NEW ZEALAND
4 AUSTRALIA
4 THAILAND
1 BOTSWANA
1 BRAZIL
1 COLOMBIA
1 ETHIOPIA
1 QATAR
1 SAUDI ARABIA
1 SINGAPORE
1 TAIWAN
1 TANZANIA
(c) Laboratories were provided with a carbon steel round bar, labelled Sample 1,
and an aluminium flat bar, labelled Sample 2. Both samples were
approximately 450 mm in length. The carbon steel round bar was
approximately 20 mm in diameter. The aluminium flat bar was approximately
40 mm in width and 6 mm in thickness.
-2-
(e) All testing, recording and reporting was to be performed in accordance with
the laboratory’s routine test methods, but testing in accordance with
AS 1391 or ISO 6892-1 were the preferred test methods.
(f) Laboratories were requested to perform the tests according to the Instructions
to Participants provided and to record the results, along with an estimate of
their measurement uncertainty (MU) for each result, on the accompanying
Results Sheet, which was distributed with the samples. Copies of these
documents appear in Appendix C.
(g) Prior to distribution, the samples were tested for homogeneity by ARL
Laboratory Services Pty Ltd. Based on the results of this testing, the
homogeneity of the samples was established (see Appendix B).
No. of Results: the total number of results for that test / sample;
Median: the middle value of the results;
Normalised IQR: the normalised interquartile range of the results;
Uncertainty of the Median: a robust estimate of the standard deviation
of the Median;
Robust CV: the robust coefficient of variation expressed as a
percentage, i.e. 100 x Normalised IQR / Median;
Minimum: the lowest laboratory result;
Maximum: the highest laboratory result; and
Range: the difference between the Maximum and Minimum.
normIQR
2 n
The ordered z-score charts in Appendix A are limited on the vertical axis to
+3.0 and -3.0, so that outliers are clearly identifiable as those laboratories
whose "bar" extends beyond the chart boundary.
-4-
For further details on the calculation and interpretation of robust z-scores and
ordered z-score charts, please see the Guide to Proficiency Testing Australia
(2019).
5. OUTLIER RESULTS
The following table summarises the results submitted by participants for the
program.
Number of Results 22 20
0.2% Proof Stress (non- Median 594.5 197.5
proportional elongation) (Rp0.2)
(MPa) Normalised IQR 25.8 19.3
Uncertainty (Median) 6.9 5.4
Number of Results 5 5
Lower Yield (ReL) Median n/a n/a
(MPa) Normalised IQR n/a n/a
Uncertainty (Median) n/a n/a
Number of Results 4 4
Upper Yield (ReH) Median n/a n/a
(MPa) Normalised IQR n/a n/a
Uncertainty (Median) n/a n/a
Number of Results 21 22
Tensile Strength (Rm) Median 630.0 229.0
(MPa) Normalised IQR 11.9 11.9
Uncertainty (Median) 3.2 3.2
Number of Results 19 20
Percentage Elongation Median 14.2 13.0
after Fracture
(A%) Normalised IQR 2.0 1.5
Uncertainty (Median) 0.6 0.4
Number of Results 19 n/a
Percentage Reduction in Median 54.0 n/a
Area after Fracture
(Z%) Normalised IQR 2.6 n/a
Uncertainty (Median) 0.8 n/a
-5-
Lower Yield
Upper Yield
Percentage Elongation
- 14, 20A, 20B
after Fracture
Percentage Reduction in
1, 17B
Area after Fracture
Notes:
1. For each test, the results for all test methods were pooled for analysis.
3. There was an insufficient number of results submitted for Lower Yield and
Upper Yield to calculate summary statistics or z-scores.
Percentage Elongation
18 out of 22 (82%) 18 out of 22 (82%)
after Fracture
Percentage Reduction in
17 out of 22 (77%)
Area after Fracture
Note that the numbers in the above table refer only to results returned that
were able to be analysed. Some laboratories attempted some of the tests but
were unable to report results for a variety of reasons (such as the sample
failing outside the gauge length, for example).
A total of 143 results were analysed in this round of the program. Of these
results, fifteen (10%) were outliers. For comparison, 5% of the results
analysed in Round 7 of the Tensile Testing of Metals program were outliers
(see Report No. 901 for more details).
-7-
For the laboratories that used the AS 1391 method for Sample 1, the median
and standard error of the 0.2% Proof Stress results was 589.0 ± 14.4 MPa.
For all methods pooled, the median and standard error of the 0.2% Proof
Stress results for Sample 1 was 594.5 ± 6.9 MPa.
For the laboratories that used the AS 1391 method for Sample 2, the median
and standard error of the 0.2% Proof Stress results was 197.0 ± 2.2 MPa. For
all methods pooled, the median and standard error of the 0.2% Proof Stress
results for Sample 2 was 197.5 ± 5.4 MPa.
The methods were pooled when analysing the results for both samples.
Four laboratories (codes 1, 13, 17B and 18) reported outlier results for
Sample 1. Two laboratories (codes 11 and 21) reported outlier results for
Sample 2. Two laboratories (codes 7 and 22) obtained absolute z-scores
between 2.0 and 3.0 for Sample 1.
The yield and proof strengths are very sensitive to loading conditions. If the
test sample is subjected to higher load rates, the 0.2% Proof Stress results
obtained may be slightly higher than normal. Another reason for high 0.2%
Proof Stress results could be that the force measurements may not have been
zeroed prior to testing and, as a result, higher strength values could be
obtained. This may be the reason for the outlier results reported by laboratory
codes 1, 13, 17B and 18, who each reported higher than normal strength
values. These participants are highly encouraged to review their testing
procedures to ensure correct loading conditions are employed during the
tensile tests.
The same can also be said for laboratory codes 11 and 21. When testing
aluminium samples, or any other material, it is imperative to check and verify
that the correct loading conditions are applied. Otherwise erroneous results
will be obtained.
The robust CV for the 0.2% Proof Stress results for Sample 1 (the round bar
sample) was 4.3%, while the robust CV for Sample 2 (the flat bar sample) was
9.8%. The last round of this program where the 0.2% Proof Stress results for
a round bar sample and a flat bar sample were analysed was Round 7. In
Round 7 of this program the robust CV obtained for the round bar sample was
2.2%, while the robust CV for the flat bar sample was 12.2%. The robust CVs
obtained for this round of the program compare well with the robust CVs
obtained in Round 7 of the program (see Report No. 901 for more details).
-8-
Only 5 laboratories determined values for the Lower Yield. Four of these
laboratories tested using ISO 6892-1, while one laboratory tested using
AS 1391 (see Appendix A7 for more details).
Only 4 laboratories determined values for the Upper Yield. Three of these
laboratories tested using ISO 6892-1, while one laboratory tested using
AS 1391 (see Appendix A7 for more details).
For the laboratories that used the AS 1391 method for Sample 1, the median
and standard error of the Tensile Strength results was 631.5 ± 9.1 MPa. For
all methods pooled, the median and standard error of the Tensile Strength
results for Sample 1 was 630.0 ± 3.2 MPa.
For the laboratories that used the AS 1391 method for Sample 2, the median
and standard error of the Tensile Strength results was 227.0 ± 2.0 MPa. For
all methods pooled, the median and standard error of the Tensile Strength
results for Sample 2 was 229.0 ± 3.2 MPa.
The methods were pooled when analysing the results for both samples.
-9-
Four laboratories (codes 1, 13, 17B and 18) reported outlier results for
Sample 1. There were no outlier results reported for Sample 2. One
laboratory (code 6) obtained an absolute z-score between 2.0 and 3.0 for
Sample 1. Four laboratories (codes 11, 14, 21 and 22) obtained absolute
z-scores between 2.0 and 3.0 for Sample 2.
The same comments regarding the high results reported by laboratory codes
1, 11, 13, 17B, 18 and 21 for 0.2% Proof Stress can also be applied to the
high results that these laboratories reported for Tensile Strength.
The robust CV for the Tensile Strength results for Sample 1 (the round bar
sample) was 1.9%, while the robust CV for Sample 2 (the flat bar sample) was
5.2%. The last round of this program where the Tensile Strength results for a
round bar sample and a flat bar sample were analysed was Round 7. In
Round 7 of this program the robust CV obtained for the round bar sample was
1.4%, while the robust CV for the flat bar sample was 2.3%. The robust CVs
obtained for this round of the program are higher than the robust CVs
obtained in Round 7 of the program (see Report No. 901 for more details).
Participants were asked to use their routine methods for testing Sample 1.
This means that the participants could employ different gauge diameters and
different gauge lengths for their tensile testing of this sample. Therefore, it
was necessary to convert the Percentage Elongation after Fracture results
submitted by the participants for Sample 1 to a proportional gauge length of
5.65 √ S0 (5.65 times the square root of the original cross-sectional area).
The results were converted to a proportional gauge length using the formula
of ISO 2566-1. These results are displayed in Appendix A5. The exact
formula used to convert the results is given on page A5.1 of Appendix A5.
For the laboratories that used the AS 1391 method for Sample 1, the median
and standard error of the Percentage Elongation after Fracture results
(converted to a proportional gauge length) was 13.6 ± 0.9%. For all methods
pooled, the median and standard error of the Percentage Elongation after
Fracture results (converted to a proportional gauge length) for Sample 1 was
14.2 ± 0.6%.
For the laboratories that used the AS 1391 method for Sample 2, the median
and standard error of the Percentage Elongation after Fracture results was
13.1 ± 0.4%. For all methods pooled, the median and standard error of the
Percentage Elongation after Fracture results for Sample 2 was 13.0 ± 0.4%.
The methods were pooled when analysing the results for both samples.
It was expected that there would be outlier results for Percentage Elongation
after Fracture for Sample 2. That is because not all participants are expected
to be familiar with aluminium material testing. The k-factor for laboratory code
14, being the lowest in this round, may suggest that the correct elongation
values were not supplied for evaluation by this laboratory. The very high
elongation values of 25%, reported by laboratory codes 20A and 20B, suggest
the use of defective measuring equipment.
The robust CV for the Percentage Elongation after Fracture results for
Sample 1 (the round bar sample) was 14.1%, while the robust CV for Sample
2 (the flat bar sample) was 11.4%. The last round of this program where the
Percentage Elongation after Fracture results for a round bar sample and a flat
bar sample were analysed was Round 7. In Round 7 of this program the
robust CV obtained for the round bar sample was 10.0%, while the robust CV
for the flat bar sample was 6.4%. The robust CVs obtained for this round of
the program are higher than the robust CVs obtained in Round 7 of the
program (see Report No. 901 for more details).
For the laboratories that used the AS 1391 method for Sample 1, the median
and standard error of the Percentage Reduction in Area after Fracture results
was 55.0 ± 1.4%. For all methods pooled, the median and standard error of
the Percentage Reduction in Area after Fracture results for Sample 1 was
54.0 ± 0.8%.
The methods were pooled when analysing the results for Sample 1.
Two laboratories (codes 1 and 17B) reported outlier results for Sample 1.
For round samples, the Percentage Reduction in Area after Fracture is one of
the most important tensile properties. If the correct cup and cone sections are
not identified, reduction areas obtained could be incorrect. Usually, the
Percentage Reduction in Area after Fracture can be calculated by measuring
the smallest diameter of the tested bar in the cup section of the broken
samples. This should be checked with the cone section of the tested sample
and should be approximately the same value. The results reported by
laboratory codes 1 and 17B suggest that these laboratories could not correctly
identify the reduced area of the tested samples.
The robust CV for the Percentage Reduction in Area after Fracture results for
Sample 1 (the round bar sample) was 4.9%. In Round 7 of this program the
robust CV obtained for the round bar sample was 4.1%. The robust CV
obtained for this round of the program compares well with the robust CV
obtained in Round 7 of the program (see Report No. 901 for more details).
Any laboratories that reported a measurement uncertainty less than two times
the uncertainty of the median may have underestimated their measurement
uncertainty.
-12-
All the participants are highly encouraged to report and use measurement
uncertainty, so that the program analysis can provide a better outlook of the
overall performance for this program. An approach, such as that described in
AS 1391 Appendix H, “An Error Budget”, to the estimation of the
measurement uncertainty in tensile testing is an example of an approach that
could be followed.
In addition to reporting results for 0.2% Proof Stress, Lower Yield, Upper
Yield, Tensile Strength, Percentage Elongation after Fracture and Percentage
Reduction in Area after Fracture, participants were also asked to report the
Tensile Specimen Gauge Diameter, Tensile Specimen Gauge Thickness,
Tensile Specimen Gauge Width, Tensile Specimen Gauge Length, Elastic
Stress or Strain Rate and Plastic Strain Rate. The details reported by each of
the participants are displayed in Appendix A7.
It is highly recommended that all of the participants report all of this necessary
information, in order to better analyse the test results. The Tensile Specimen
Diameter and Tensile Specimen Gauge Length were required in order to
convert the Percentage Elongation after Fracture results to a proportional
gauge length for Sample 1. The Tensile Specimen Gauge Thickness, Tensile
Specimen Gauge Width and Tensile Specimen Gauge Length can be used to
check if the participants machined Sample 2 according to the Instructions to
Participants. The loading rates (both stress and strain rates) can be reviewed
for any abnormal test results. The loading rates are particularly important
when comparing test results for elongation and yield strength, as the loading
rates can affect these properties.
The aim of this round of the program was to determine the capabilities and
proficiency of the participating laboratories in comparing and understanding
the role of specimen geometry and material type on the tensile properties of
the materials.
All the participating laboratories that reported results that have an absolute
z-score of more than 2.0 are highly encouraged to use these proficiency
testing programs to evaluate their testing methods, loading conditions,
preparation procedures, equipment management and training details. It is
expected of each participant to analyse their own proficiency testing results
and evaluate any trends.
7. REFERENCES
6. ASTM B557M – Standard Test Methods for Tension Testing Wrought and
Cast Aluminum and Magnesium-Alloy Products (Metric).
Summary of Results
Section A1
Summary Statistics
Notes:
1. § denotes an outlier (i.e. |z-score| ≥ 3.0).
2. Laboratory code 5 could not report a result for 0.2% Proof Stress for Sample 2
because their extensometer failed.
3. The samples for laboratory code 12 were machined by another laboratory.
17B
13
18
1
3
7
2
6
Robust Z-Score
14
8
21
15
5
0
2
20A
17A
20B
12
19
9
-1
4
24
11
-2
22
-3
Laboratory Code
21
3
2
1
Robust Z-Score
16
1
23
6
20B
24
18
12
0
17A
7
19
13
-1
20A
9
8
14
22
-2
-3
11
Laboratory Code
Section A2
Lower Yield
A2.1
Notes:
1. There was an insufficient number of results submitted for Lower Yield to
calculate summary statistics or z-scores.
Section A3
Upper Yield
A3.1
Notes:
1. There was an insufficient number of results submitted for Upper Yield to
calculate summary statistics or z-scores.
Section A4
Tensile Strength
A4.1
Summary Statistics
Notes:
1. § denotes an outlier (i.e. |z-score| ≥ 3.0).
2. Laboratory code 7 machined Sample 1 at full section. The sample failed within
the jaws and so the Tensile Strength could not be reported for Sample 1.
3. The samples for laboratory code 12 were machined by another laboratory.
17B
13
18
1
3
6
2
20B
11
8
Robust Z-Score
20A
5
15
2
0
14
12
22
17A
21
-1
4
24
19
9
-2
-3
Laboratory Code
21
3
11
4
2
Robust Z-Score
20B
16
23
1
17B
6
24
12
5
0
17A
7
18
19
-1
13
9
20A
8
-2
14
22
-3
Laboratory Code
Section A5
Percentage Elongation
after Fracture
A5.1
Summary Statistics
Notes:
1. § denotes an outlier (i.e. |z-score| ≥ 3.0).
2. * The sample failed outside of gauge length.
3. Laboratory code 7 machined Sample 1 at full section. The sample failed within
the jaws and so the Percentage Elongation after Fracture could not be reported
for Sample 1.
4. The samples for laboratory code 12 were machined by another laboratory.
14
21
11
2
15
24
1
Robust Z-Score
5
19
22
4
17A
0
2
9
8
6
13
-1
1
12
18
17B
-2
-3
Laboratory Code
20B
14
2
5
18
1
Robust Z-Score
11
9
21
17B
17A
19
24
7
0
13
16
-1
23
12
6
8
-2
-3
4
Laboratory Code
Section A6
Summary Statistics
Statistic Sample 1
Number of Results 19
Median 54.0
Normalised IQR 2.6
Uncertainty (Median) 0.8
Robust CV 4.9%
Minimum 46
Maximum 66
Range 20
A6.2
Notes:
1. § denotes an outlier (i.e. |z-score| ≥ 3.0).
2. Laboratory code 7 machined Sample 1 at full section. The sample failed within
the jaws and so the Percentage Reduction in Area after Fracture could not be
reported for Sample 1.
3. The samples for laboratory code 12 were machined by another laboratory.
4. The Percentage Reduction in Area after Fracture was only required to be
reported for Sample 1.
17B
3
5
9
22
1
Robust Z-Score
17A
11
19
2
6
14
4
0
8
12
-1
20A
13
24
18
20B
-2
-3
1
Laboratory Code
Section A7
Method Information
Lab
0.2% Proof Stress Lower Yield Upper Yield
Code
1 ISO 6892-1: 2019 ISO 6892-1: 2019 ISO 6892-1: 2019
2 BS EN ISO 6892-1: 2016 - -
4 BS 6892-1 BS 6892-1 -
5 AS 1391 - -
6 AS 1391 - -
7 AS 1391 - -
8 ISO 6892-1 - -
9 AS 1391 - 2007 (R2017) - -
11 ISO 6892-1: 2016 - -
12 AS 1391: 2007 - -
13 ISO 6892 - -
14 - - -
15 - - -
16 JIS Z 2241: 2011 - -
17A AS 1391 - -
17B AS 1391 - -
18 AS 1391 AS 1391 AS 1391
19 AS 1391 - -
20A AS 1391 - 2007 (R2017) - -
20B AS 1391 - 2007 (R2017) - -
21 ISO 6892-1 ISO 6892-1 ISO 6892-1
22 ISO 6892-1: 2019 ISO 6892-1: 2019 ISO 6892-1: 2019
23 ASTM B557M - -
24 ISO 6892-1 (2016) - -
Note:
Method information for Lower Yield and Upper Yield has only been reported for
those laboratories that determined values for these tests.
A7.2
Method Information
Lab Percentage Elongation Percentage Reduction in
Tensile Strength
Code after Fracture Area after Fracture
1 ISO 6892-1: 2019 ISO 6892-1: 2019 ISO 6892-1: 2019
2 BS EN ISO 6892-1: 2016 BS EN ISO 6892-1: 2016 BS EN ISO 6892-1: 2016
4 BS 6892-1 BS 6892-1 BS 6892-1
5 AS 1391 AS 1391 AS 1391
6 AS 1391 AS 1391 AS 1391
7 AS 1391 AS 1391 -
8 ISO 6892-1 ISO 6892-1 ISO 6892-1
9 AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017)
11 ISO 6892-1: 2016 ISO 6892-1: 2016 ISO 6892-1: 2016
12 AS 1391: 2007 AS 1391: 2007 AS 1391: 2007
13 ISO 6892 ISO 6892 ISO 6892
14 - - -
15 - - -
16 JIS Z 2241: 2011 JIS Z 2241: 2011 -
17A AS 1391 AS 1391 AS 1391
17B AS 1391 AS 1391 AS 1391
18 AS 1391 AS 1391 AS 1391
19 AS 1391 AS 1391 AS 1391
20A AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017)
20B AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017) AS 1391 - 2007 (R2017)
21 ISO 6892-1 ISO 6892-1 -
22 ISO 6892-1: 2019 ISO 6892-1: 2019 ISO 6892-1: 2019
23 ASTM B557M ASTM B557M -
24 ISO 6892-1 (2016) ISO 6892-1 (2016) ISO 6892-1 (2016)
A7.3
Homogeneity
Testing
B1.1
HOMOGENEITY TESTING
Before the test pieces were distributed to participants, eight specimens from each
sample were selected at random and tested by ARL Laboratory Services Pty Ltd. This
was done to assess the variability of the samples to be used in the program. The results
of this testing appear in the following tables.
Analysis of this testing data indicated that the samples were sufficiently homogeneous for
the program and, therefore, any participant results identified as outliers cannot be
attributed to sample variability.
APPENDIX C
Instructions to Participants
and
Results Sheet
C1.1
Instructions To Participants
To ensure that the results of this program can be analysed correctly, participants are
asked to note carefully:
1) The samples for this tensile testing program comprise of a carbon steel round
bar, labelled 1-x for Sample 1, and an aluminium flat bar, labelled 2-x for
Sample 2.
3) Tests may commence as soon as samples are received. The carbon steel
sample (Sample 1) is to be treated in the same manner as your routinely tested
samples. For the aluminium sample (Sample 2), a Tensile Specimen Gauge
Length of 80 mm, Tensile Specimen Gauge Width of 20 mm, Parallel Length of
90 mm and Minimum Radius of Transition of 20 mm must be used (please see
the diagram on page 3 of these Instructions).
5) Report only one result per sample, based on the determination for each
property. For each determination, results are to be reported to the accuracy
and in the units indicated on the Results Sheet.
6) The method of testing used should also be reported on the Results Sheet (e.g.
AS 1391, ISO 6892-1, etc.)
9) For this program, your laboratory has been allocated the code number on the
attached Results Sheet. All reference to your laboratory in reports associated
with this program will be via this code number, ensuring the confidentiality of
your results.
10) Return the Results Sheet, either by mail, email or facsimile, to:
Mark Bunt
Proficiency Testing Australia
PO Box 7507
Silverwater NSW 2128
AUSTRALIA
Telephone: + 61 2 9736 8397 (1300 782 867)
Fax: +61 2 9743 6664
Email: mbunt@pta.asn.au
All results should arrive at the above address by no later than Wednesday 20
May 2020. Results reported later than this date may not be analysed in the
final report.
Where possible, please also report the values for the following: