Hardness Testing of Metals Proficiency Testing Program Round 9

REPORT NO. 963

Hardness Testing of Metals

Proficiency Testing Program

Round 9

March 2016

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.

© COPYRIGHT PROFICIENCY TESTING AUSTRALIA 2016

PO Box 7507 Silverwater NSW 2128 AUSTRALIA

CONTENTS

1. FOREWORD 1

2. FEATURES OF THE PROGRAM 1

3. FORMAT OF THE APPENDICES 2

4. STATISTICAL DESIGN OF THE PROGRAM 2

5. OUTLIER RESULTS 2

Table A: Summary Statistics for All Tests 3

Table B: Summary of Statistical Outliers 4

6. PTA AND TECHNICAL ADVISER'S COMMENTS 4

7. REFERENCES 7

APPENDICES

APPENDIX A

Summary of Results

Brinell Hardness A1.1

Vickers Hardness A2.1

Rockwell Hardness A3.1

Methods Used and Surface Preparation A4.1

APPENDIX B

Homogeneity Testing B1.1

APPENDIX C

Instructions to Participants C1.1

Results Sheet C2.1

-1-

1. FOREWORD This report summarises the results of a proficiency testing program on the

hardness properties of metals. It constitutes the ninth round of an ongoing series of programs.

Proficiency Testing Australia (PTA) conducted the testing program in December

2015 / January 2016. The aim of the program was to assess laboratories' ability to competently perform the nominated tests.

The Program Coordinator was Dr M Bunt. The Technical Adviser was

Mr S Sameem, ARL Laboratory Services Pty Ltd. This report was authorised by Mr P Briggs, PTA General Manager.

2. FEATURES OF THE PROGRAM (a) A total of 14 laboratories participated in the program, all of which returned results

for inclusion in the final report. Laboratories from the following countries received samples:

11 AUSTRALIA 1 COLOMBIA 1 ENGLAND 1 SAUDI ARABIA

To ensure confidential treatment of results, each laboratory was allocated a unique random code number. Reference to each laboratory in this report is by its code number. Please note that one laboratory reported more than one set of results and, therefore, this laboratory’s code number (with letter) could appear several times in the same data set.

(b) The results reported by participants are presented in Appendix A.

(c) Each laboratory was provided with a steel sample, approximately 60 mm in diameter and 30 mm thick. The sample was to be tested for Brinell, Vickers and either Rockwell A or Rockwell B hardness testing.

(d) 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.

(e) 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).

-2-

3. FORMAT OF THE APPENDICES (a) Appendix A is divided into four sections (A1-A4).

Sections A1-A3 contain the analysis of results reported by laboratories for Brinell, Vickers and Rockwell hardness. These sections contain: i) a table of results reported by laboratories for each test, with estimates of

their MUs and calculated z-scores; ii) a listing of the summary statistics; and iii) ordered z-score charts.

Section A4 contains information on the methods used by the participants and the surface preparations they performed.

(b) Appendix B contains details of the homogeneity testing. (c) Appendix C contains copies of the Instructions to Participants and Results Sheet. 4. STATISTICAL DESIGN OF THE PROGRAM

For each statistically analysed test, robust statistical procedures were used to generate the z-scores and summary statistics for each hardness determination - number of results, median, normalised interquartile range (IQR), uncertainty of the median, robust coefficient of variation (CV), minimum, maximum and range. For each participant, the robust z-score calculated for each hardness result was based on the average hardness value reported for each test.

5. OUTLIER RESULTS

Robust z-scores have been used to assess each laboratory’s testing performance. When calculated from single results, z-scores are used to detect excessively high or excessively low results in comparison to the consensus value (the median). Any result with an absolute z-score greater than or equal to 3.0 (i.e. ≤ -3.0 or ≥ 3.0) is classified as an outlier. For further details on the calculation and interpretation of robust z-scores, please see reference [1].

-3-

The following table summarises the results submitted by participants for the program.

Table A: Summary Statistics for All Tests

Test Summary Statistics Average Result

Brinell Hardness

(HBW)

Number of Results 11

Median 209.0

Normalised IQR 4.8

Uncertainty (Median) 1.8

Vickers Hardness

(HV)


Median 215.5

Normalised IQR 5.0


Rockwell B Hardness

(HRB)


Median 95.70

Normalised IQR 0.79


Notes: 1. For each test, the results for all test methods were pooled for analysis. 2. Summary statistics were calculated for the average hardness value reported

for each test.

3. The uncertainty of the median was calculated as: n

normIQR

2.

4. All Rockwell Hardness results were reported as Rockwell B Hardness.

-4-

Table B: Summary of Statistical Outliers (By laboratory code number)

Test Outliers (Laboratory Code No.)

Brinell Hardness -

Vickers Hardness -

Rockwell B Hardness 3

6. PTA AND TECHNICAL ADVISER’S COMMENTS

Consensus values (medians), derived from participants’ results, are used as the assigned values in this program. These values are not metrologically traceable to an external reference. The summary statistics, uncertainties of the assigned values and outliers, for each of the tests, are reported in Tables A and B above. Complete details of the statistical analyses appear in Appendix A.

6.1 Return rate

All of the 14 laboratories that participated in the program returned results. Of these 14 laboratories, 9 (64%) submitted results for all three tests. The return rate for all tests is as follows:

Brinell Hardness 11 out of 14 79%

Vickers Hardness 11 out of 14 79%

Rockwell B Hardness 11 out of 14 79%

6.2 Performance summary

Statistical outliers were reported by one of the 14 laboratories (7%) that returned results in this round of the program. For comparison, 7% of the participants reported outlier results in Round 8 of this program (see Report No. 916 for more details). A total of 35 results were analysed in this program. Of these results, one (3%) was an outlier result. For comparison, 3% of the results analysed in Round 8 of this program were outlier results (see Report No. 916 for more details).

-5-

6.3 Brinell Hardness

A total of 11 laboratories tested the sample for Brinell hardness. Of these laboratories, seven tested using the AS 1816 method. Another laboratory indicated that they used an Australian Standard method, but did not identify if this was AS 1816. One laboratory used both AS 1816 and ISO 6506. Two laboratories tested using the ASTM E10 method (see Appendix A4 for more details). The methods were pooled when analysing the results. For all methods pooled, the median and standard error of the Brinell hardness results was 209.0 ± 1.8 HBW. The CV for the Brinell hardness results for this round was 2.3%. This is higher than the CV of 0.9%, obtained for the Brinell hardness results in Round 8 of this program (see Report No. 916). There were no outliers reported for Brinell hardness. Seven laboratories reported measurement uncertainties associated with their Brinell hardness test results in this round.

6.4 Vickers Hardness

Of the 11 laboratories that tested the sample for Vickers hardness, seven tested using the AS 1817 method. Another laboratory indicated that they used an Australian Standard method, but did not identify if this was AS 1817. One laboratory used both AS 1817 and ISO 6507. Two laboratories tested using the ASTM E384 method, including one laboratory that submitted two sets of results (see Appendix A4 for more details). The methods were pooled when analysing the results. For all methods pooled, the median and standard error of the Vickers hardness results was 215.5 ± 1.8 HV. The CV for the Vickers hardness results for this round was 2.3%. This is lower than the CV of 4.2%, obtained for the Vickers hardness results in Round 8 of this program (see Report No. 916). There were no outliers reported for Vickers hardness. Six laboratories reported measurement uncertainties associated with their Vickers hardness test results in this round, including one laboratory that reported measurement uncertainties for their two sets of results.

-6-

6.5 Rockwell Hardness

A total of 11 laboratories tested the sample for Rockwell hardness. Although laboratories could report the Rockwell hardness of their sample as either Rockwell A or Rockwell B hardness, all of these 11 laboratories reported their results as Rockwell B hardness. Of these laboratories, seven tested using the AS 1815 method. Another laboratory indicated that they used an Australian Standard method, but did not identify if this was AS 1815. Two laboratories tested using the ASTM E18 method, including one laboratory that submitted two sets of results. One laboratory did not specify their method of testing (see Appendix A4 for more details). The methods were pooled when analysing the results. For all methods pooled, the median and standard error of the Rockwell B hardness results was 95.70 ± 0.28 HRB. The CV for the Rockwell B hardness results for this round was 0.8%. This is lower than the CV of 1.7%, obtained for the Rockwell B hardness results in Round 8 of this program (see Report No. 916). Laboratory 3 reported an outlier result for Rockwell B hardness. Two other laboratories (codes 4 and 7A) obtained absolute z-scores between 2.0 and 3.0. A good hardness testing procedure should consider the microstructural effects and spacing of hardness impressions. Amongst other factors, this indentation spacing depends mainly on the type of loading used. For instance, the larger the loading applied, the greater the spacing required between hardness impressions. This is to avoid testing the affected microstructure due to local deformation of previous impressions that may produce inaccurate results. This could explain the Rockwell hardness results for laboratory 3, whose average of 99 HRB, compared to the median of 95.7 HRB, resulted in an outlier (z-score of 4.20). For Vickers hardness testing, laboratory 3 indicated that a loading of 50 kg was used. That would have created a larger local deformation. One of the reasons for a higher Rockwell result for laboratory 3 may be that the spacing for their impressions was not adequate, giving the laboratory a higher reading. For a 50 kg load, a minimum of 5 mm - 10 mm of spacing may get accurate results. Another way laboratory 3 could improve their results would be to compare their hardness values. There are numerous resources online, for example ASTM E140-12be1, that can be used to convert results from different hardness methods. However, if the deformation is still large, then the surface can be milled down by 5 mm - 10 mm and the hardness preparation and testing can be redone. All the results should be analysed before being submitted. Seven laboratories reported measurement uncertainties associated with their Rockwell B hardness test results in this round, including one laboratory that reported measurement uncertainties for their two sets of results.

-7-

6.6 Measurement Uncertainty

There seems to have been some confusion on reporting measurement uncertainty for some participants. The variation between the estimates of measurement uncertainty range from not reporting at all, to some reporting percentages, while others reported numerical values. All the participants are highly encouraged to report and use measurement uncertainty, so that the program analysis could be simpler and provide a better outlook of the overall performance of these testing programs.

7. REFERENCES

1. Guide to Proficiency Testing Australia (2014). (This document is located on the PTA website at www.pta.asn.au under Programs / Documents).

2. AS 1815.1: 2007 Metallic materials – Rockwell hardness test – Test method

(scales A, B, C, D, E, F, G, H, K, N, T). 3. AS 1816.1: 2007 Metallic materials – Brinell hardness test – Test method

(ISO 6506-1: 2005, MOD). 4. AS 1817.1: 2003 Metallic materials – Vickers hardness test – Test method

(ISO 6507-1: 1997, MOD). 5. ISO 6506-1: 2014 Metallic materials – Brinell hardness test – Part 1: Test

method. 6. ISO 6507-1: 2005 Metallic materials – Vickers hardness test – Part 1: Test

method. 7. ISO 6508-1: 2015 Metallic materials – Rockwell hardness test – Part 1: Test

method.

8. ASTM E10 – Standard Test Method for Brinell Hardness of Metallic Materials.

9. ASTM E18 – Standard Test Methods for Rockwell Hardness of Metallic Materials.

10. ASTM E384 – Standard Test Method for Knoop and Vickers Hardness of Materials.

11. ASTM E140-12be1 – Standard Hardness Conversion Tables for Metals Relationship Among Brinell Hardness, Vickers Hardness, Rockwell Hardness, Superficial Hardness, Knoop Hardness, Scleroscope Hardness, and Leeb Hardness.

http://www.pta.asn.au/

APPENDIX A

Summary of Results

Section A1

Brinell Hardness

A1.1

Brinell Hardness (HBW) – Results and Z-Scores

Lab Code

Temp. (oC)

Test 1 Test 2 Test 3 Average MU (±) Z-Score

1 23 212 210 214 212 - 0.62

3 22.2 209 209 210 209 12 0.00

4 20 212 212 212 212 - 0.62

5 20 215 216 215 215 0.37% 1.25

7A 21 208 208 208 208 3.8% -0.21

8 20 203.0 204.0 204.3 203.8 5% -1.08

9 23 222 212 214 216 - 1.45

10 20 217 217 212 215 0.010 1.25

11 22 209 209 209 209 4 0.00

13 23 198 204 205 202.3 5 -1.39

14 24 206 206 206 206 - -0.62

Summary Statistics

Statistic Average Result


Median 209.0

Normalised IQR 4.8


Robust CV 2.3%

Minimum 202

Maximum 216

Range 14

Note: 1. Summary statistics and z-scores have been calculated for the average results

reported.

A1.2

13

8

14

7A

3 11

1 4

5 10

9

-3

-2

-1

0

1

2

3

Robust

Z-S

core

Laboratory Code

Brinell Hardness (HBW)

Section A2

Vickers Hardness

A2.1

Vickers Hardness (HV) – Results and Z-Scores

Lab Code

Load Temp. (oC)


1 10 23 218.33 219.06 218.33 218.57 - 0.62

2 10 23 209 207 206 207 7 -1.71

3 50 22.0 222 217 219 219 16 0.71

4 10 28 208 210 215 211 - -0.91

7A 30 21 215 214 213 214 7.4% -0.30

7B 10 22 211 212 213 212 2.3% -0.71

8 20 20 217.2 218.4 218.2 217.9 5% 0.48

9 30 23 219 216 217 217 - 0.30

11 30 22 222 219 213 218 6 0.50

12 10 23.5 221 221 221 221 - 1.11

13 10 23 211 213 211 211.6 5 -0.79

14 10 24 205 206 206 206 - -1.91

Summary Statistics



Median 215.5

Normalised IQR 5.0


Robust CV 2.3%

Minimum 206

Maximum 221

Range 15

Note:

1. Summary statistics and z-scores have been calculated for the average results

reported.

A2.2

14

2

4 13 7B

7A

9

8 11 1 3

12

-3

-2

-1

0

1

2

3

Robust

Z-S

core

Laboratory Code

Vickers Hardness (HV)

Section A3

Rockwell Hardness

A3.1

Rockwell B Hardness (HRB) – Results and Z-Scores

Lab Code

Temp. (oC)


1 23 96.3 96.5 96.1 96.3 - 0.76

3 22 99 97 100 99 2 4.20 §

4 20 93.4 94.8 93.9 94.0 - -2.16

6 21.0 96.27 96.20 96.07 96.18 0.18 0.61

7A 21 94.0 94.0 94.4 94.1 1.8% -2.04

7B 22 96.7 94.9 95.1 95.6 1.2% -0.13

8 20 96.10 96.12 96.14 96.12 5% 0.53

9 23 95.0 96.2 96.3 95.8 - 0.13

11 22 95.8 95.5 94.1 95.1 2 -0.76

12 23.5 96 95.5 96 95.8 0.72 0.13

13 23 95.5 95.5 96 95.6 2 -0.13

14 24 94 95 95 95 - -0.89

Summary Statistics



Median 95.70

Normalised IQR 0.79


Robust CV 0.8%

Minimum 94.0

Maximum 99.0

Range 5.0

Notes:

1. Summary statistics and z-scores have been calculated for the average results

reported.

2. All Rockwell Hardness results were reported as Rockwell B Hardness.

A3.2

4 7A

14 1

1

7B 13

9 12

8

6

1

3

-3

-2

-1

0

1

2

3

Robust

Z-S

core

Laboratory Code

Rockwell B Hardness (HRB)

Section A4

Methods Used and Surface Preparation

A4.1

Methods Used

Lab Code

Brinell Hardness Vickers Hardness Rockwell B Hardness

1 ASTM E10 ASTM E384 ASTM E18

2 - AS 1817.1: 2003, ISO 6507.1: 2005 -

3 AS 1816.1 AS 1817.1 AS 1815.1

4 AS 1816.1 AS 1817.1 AS 1815.1

5 AS 1816.1 - -

6 - - -

7A ASTM E10 ASTM E384 ASTM E18

7B - ASTM E384 ASTM E18

8 AS 1816.1: 2007 AS 1817.1: 2003 AS 1815.1: 2007

9 AS 1816 AS 1817 AS 1815

10 AS 1816.1, ISO 6506.1 - -

11 AS 1816.1 AS 1817.1 AS 1815.1

12 - AS 1817.2: 2002 AS 1815.3: 2002

13 AS AS AS

14 AS 1816.1 AS 1817.1 AS 1815.1

A4.2

Surface Preparation

Lab Code

Preparation Details

1 Milling and then finish grinding.

2 Abraided in sequence from 80 to 1200 grit SiC paper and polished to 1 µm.

3 Grind and polish to 1 micron finish.

4 Snow ground blue surface. For Vickers test, the blue surface was ground to 1200 grit.

5 Turned then fine grind to 0.2 Ra.

6 Milling, 5/10 rough, with sandpaper 120, 320, 400, 600 and 1000. Polished monocrystalline diamond suspension. ASTM E3.

7A Ground to 1200 grit.

7B Ground to 1200 grit.

8 The sample was polished to 2000 grit abrasive paper and then etched.

9 Using a lathe - faced both sides of test piece to make them parallel and flat. Finished blue test surface with wet and dry paper final diamond paste finish to 3 µm.

10 Site was prepared using a small flap disc and belt sander to clear all machining marks and paint.

11 Metallographic wet abrasion P 120, P 240, P 320 P 600 → P 1200 final finish.

12 The sample was wet ground and mechanically polished down to a 1 micron surface finish.

13 Linished and polished to P 1200.

14 Ground at grit 80 until flat. Polished to 1200 grit surface finish.

APPENDIX B

Homogeneity Testing

B1.1

HOMOGENEITY TESTING Before the samples were distributed to participants, eight randomly selected samples were tested for homogeneity by ARL Laboratory Services Pty Ltd. The results of the homogeneity testing are displayed below:

Homogeneity Testing Results

Brinell HBW 10/3000

Sample No. Test 1 Test 2 Test 3 Test 4 Test 5 Average

3 202 201 202 203 202 202.0

13 202 201 203 202 201 201.8

17 203 204 202 203 201 202.6

18 202 203 202 203 202 202.4

22 202 202 203 203 202 202.4

27 205 205 204 205 202 204.2

31 205 204 205 204 202 204.0

39 205 204 204 204 202 203.8

Vickers HV 10


3 220 220 220 220 218 219.6

13 218 219 219 219 218 218.6

17 216 220 218 222 216 218.4

18 218 220 218 218 216 218.0

22 220 219 220 221 217 219.4

27 217 220 220 220 217 218.8

31 217 219 220 221 218 219.0

39 219 220 220 217 218 218.8

B1.2

Rockwell HRB


3 96.7 96.4 96.7 97.0 96.3 96.62

13 94.9 96.1 96.9 96.3 94.1 95.66

17 96.1 96.4 95.6 96.0 95.3 95.88

18 96.2 96.7 96.0 94.4 95.5 95.76

22 96.3 96.1 96.7 96.0 95.6 96.14

27 96.0 95.5 96.3 96.5 96.3 96.12

31 96.0 95.8 96.7 96.8 95.2 96.10

39 96.3 96.6 96.0 96.6 95.9 96.28

Analysis of the homogeneity 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

Hardness Testing of Metals, Round 9 December 2015 Page 1 of 2

PROFICIENCY TESTING AUSTRALIA

Hardness Testing Of Metals Proficiency Testing Program Round 9, December 2015

Instructions to Participants

To ensure that the results of this program can be analysed correctly, participants are asked to adhere carefully to these instructions. 1) The sample for this hardness testing program consists of a steel sample, approximately 60mm

in diameter and 30mm thick. The sample has been painted blue on one side and yellow on the other side, and has a number labelled on the circumference.

2) The sample is to be tested for Brinell, Vickers and Rockwell Hardness (recommended Scale A or B). Participants may wish to improve the surface to provide a better testing surface. The side painted blue is the side that should be prepared and tested.

3) The sample should be treated as a routine laboratory sample. All testing, recording and reporting is to be performed in accordance with your routine test methods.

4) Please use the attached result sheet to record and report your results to Proficiency Testing Australia. Please also report the method used for testing (e.g. AS 1816.1, ISO 6506.1, etc. for Brinell hardness testing, AS 1817.1, ISO 6507.1, etc. for Vickers hardness testing, AS 1815.1, ISO 6508.1, etc. for Rockwell hardness testing).

5) Do not discard the hardness test sample until you have received the final report. You may be

asked to carry out a retest or to return the sample to Proficiency Testing Australia for retesting at the laboratory that performed the homogeneity tests.

6) 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, thus ensuring the confidentiality of your results.

7) Laboratories are also requested to calculate and report an estimate of uncertainty of

measurement for each reported measurement result. All estimates of uncertainty of

measurement must be given as a 95% confidence interval (coverage factor k 2).

8) Return the Results Sheet, either by mail, email or facsimile, to:

All results should arrive at the above address by no later than Friday 22 January 2016.

Results reported later than this date may not be analysed in the final report.

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: [email protected]

C2.1

Hardness Testing of Metals, Round 9 December 2015 Page 2 of 2

PROFICIENCY TESTING AUSTRALIA

Hardness Testing Of Metals Proficiency Testing Program

Round 9, December 2015

RESULTS SHEET

Laboratory Code:

Sample I.D.

Scale Report

to nearest

Test Temp

ºC

Results Standard (AS, ISO,

etc.) Test 1 Test 2 Test 3 Average MU (±)

Brinell

( / / ) (mm/kg/s)

1 BHN

HV ( )

Insert load used 1 HV

Rockwell ( )

Insert Scale 0.1 HR (A / B)

Did you carry out surface preparation on the obverse surfaces of the samples for the hardness tests? Yes / No

If Yes, please give details of preparation.

…….…………………………………………………………………………………………………………………

…….…………………………………………………………………………………………………………………

…….…………………………………………………………………………………………………………………

…….…………………………………………………………………………………………………………………

…….…………………………………………………………………………………………………………………

…….…………………………………………………………………………………………………………………

Print Name: __________________________

Signature: ___________________________

Date: _______________________________

-----End of Report-----