Torque Key Comparison CCM.T-K1.3 Measurand Torque: 0 N m, 500 N m, 1000 N m Final Report March 9, 2015 Pilot Institute: Contact Person: E-mail: Co-author: Korea Research Institute of Standards and Science, Republic of Korea Min-Seok Kim minsk@kriss.re.kr K. M. Khaled (National Institute of Standards, Egypt)
1. Introduction The comparison on torque in the range of 0 N m, 20 N m, 50 N m, 500 N m, 1000 N m, and 2000 N m was organized and conducted between two laboratories, Korea Research Institute of Science and Standards (KRISS) and National Institute of Standard (NIS) in Egypt. The purpose of this comparison is to provide a link to CCM.T-K1 of the national torque standards in Egypt so that NIS can achieve the degree of equivalence of the CMCs of his torque standards. KRISS, who joined CCM.T-K1 key comparison in 2005, plays a role of a link laboratory. The results of this comparison are not used to determine the key comparison reference values (KCRVs) for CCM.T-K1, but rather the KCRVs of CCM.T-K1 will be used as reference values as usual in a follow-up comparison. Details of this comparison will be described in two sections. The results of comparison in the range of 0 N m, 500 N m and 1000 N m, which will be linked to those of CCM.T-K1, are described in Section 2. The results of additional comparison in the range of 0 N m, 20 N m, 50 N m, 1 kn m and 2 kn m are described in the Appendix. NIS used a different torque standard machine when conducting the additional comparison experiments. 2. Comparison on measurand torque at 0 N m, 500 N m and 1000 N m 2.1 Participants details The participating laboratories are KRISS (pilot) and NIS (Egypt). Their torque standard machines (TSMs) and contact details are listed in Table 1 and 2, respectively. Table 1: Participated institutes and torque standard machines Torque Standard Machine Institute Capacity in kn m Type of applied torque Note KRISS 2 Deadweight 0.005 Link Lab. NIS 1 Comparator (Horizontal) 0.06 Reference transducers traceable to KRISS
Table 2: Contact details KRISS Min-Seok Kim 267 Gajeong-Ro, Yuseong-Gu, Daejeon 305-340, Republic of Korea Tel: +82-42-868-5242 Fax: +82-42-868-5249 e-mail: minsk@kriss.re.kr NIS K. M. Khaled Tersa street, Haram, Giza, P.O. Box 136 Giza Code 11221, Egypt Tel: +202-3387-9243 Fax: +202-3387-9244 e-mail: khaled_fmmd_nis@yahoo.com 2.2 Comparison Protocol The protocol of this CCM.T-K1.3 basically is similar to that of CCM.T-K1 key comparison [1]. The differences are as follows; 1) One travelling standard with 1000 N m capacity was used to compare 0 N m, 500 N m, and 1000 N m torques due to practical limitation. 2) We strictly followed the measurement sequence of CCM.T-K1 except the time interval. The time interval was reduced to 5 minutes in this comparison from 6 minutes in CCM.T-K1 in order to complete the measurement sequence more quickly while minimizing the creep effect of the transfer standard. 3) The same bridge amplifier was used for indications of all travelling standards for both laboratories. Thus, a bridge calibration unit was not used to calibrate the amplifier. We expect that the correction due to the long-term drift of the amplifier itself can be ignored. The measurement sequence is listed in Table 3 and depicted in Figure 1.
Table 3: Measurement sequence of the Torque KC CCM.T-K1.3 Loadings Position Steps 3 initial pre-loadings 0 3 measurement cycles 60 120 180 240 300 0 N m, 500 N m, 360 1000 N m 420 480 540 600 660 720
0 0 60 120 720 three preloadings three meas. cycles preloading preloading meas. cycles 1000 Torque / (N m) 500 0-10 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 Time / min Figure 1: Diagram of the measurement sequence of the torque KC. 2.3 Traveling standard A torque transducer with a capacity of 1000 N m (belonging to NIS) was used. Its details are listed as follows: maker: Hottinger Baldwin Messtechnik GmbH, Germany type: TN serial number: #014440034/1 kn m adaptation: 50 h7 80 shaft end on both sides The details of the bridge amplifier(belonging to NIS) used in this comparison and its settings are as follows: maker: Hottinger Baldwin Messtechnik GmbH, Germany type: DMP 40
serial number: 022620004 filter: 0.2 Hz Bessel signal reading: Absolute Measuring range: 2.5 mv/v Excitation voltage: 5 V Auto calibration: AcalOFF 2.4 Comparison formation The measurement was carried out one time for each participant due to the practical limitation. To reduce the effect of long-term stability of the travelling standard, the time interval between two successive measurements was set to approximately one and half month, which includes transportation time, thermal stabilization time of the travelling standard at each laboratory. 2.5 Measurement results The measurement result is the mean deflection calculated from 12 original readings measured in 12 orientations (60 to 360 for two rotations) for each of the two torque steps and the two directions. The measurement has to be calculated for the mean deflection measured with the transducer and the DMP 40. No corrections for amplifier, temperature, air humidity or creep were applied in those results. We expect that these effects could be negligible compared to the relative of 0.06 % of NIS TSM based on the report of CCM.T-K1 [1]. The was calculated according to the protocol of CCM.T-K1 and the same worksheets were used to report the results. The measurement results are listed in Table 4.
Table 4: Reported deflections and relative uncertainties (k = 2) Institute Date Deflection Clockwise torque 500 N m 1000 N m Deflection KRISS 2010-11-09 0.877832 0.0058 1.755772 0.0058 NIS 2010-12-23 0.877753 0.06 1.755796 0.06 Institute Date Deflection Counterclockwise torque 500 N m 1000 N m Deflection KRISS 2010-11-09-0.877804 0.0057-1.755718 0.0057 NIS 2010-12-24-0.877774 0.06-1.755793 0.06 2.6 Comparison results In order to test the equivalence of reported data among participants, the analysis was usually performed according to [2], but in this comparison, modifications are required due to correlation between reported values from two NMIs. 2.6.1 Deviation of the measured values Deviation of a measurement result of NIS (x NIS ) from that of KRISS (x KRI ) can be calculated as follows, N-K = NIS KRI (1) The standard of the deviation, N-K, should be evaluated in consideration of correlation between the measurement values of NIS and KRISS. The N-K can be calculated from ( N-K ) = ( KRI ) + ( NIS ) 2 ( KRI, NIS ) ( KRI ) ( NIS ) (2)
The correlation coefficient ( KRI, NIS ) is not zero because the reference torque transducer used in the NIS comparative-type torque calibration machine was calibrated using the KRISS torque standard. The correlation coefficient (, ) can be estimated approximately by [3] ( KRI, NIS ) ( KRI ) NIS ( NIS ) KRI (3) where, KRI and NIS are a change in KRI and NIS, respectively. All values of the correlation coefficient in our comparison are close to 0.1 because KRI and NIS are almost same (i.e., deflections of KRISS and NIS are almost same as seen in Table 4) and the ratio of uncertainties, ( KRI )/ ( NIS ), is approximately 0.1. The calculated results are tabulated in Table 5 and graphically shown in Figure 2 and Figure 3. Table 5: Deviations of the measured values of NIS from the pilot, KRISS Deviation from the pilot Torque N-K in nv/v N-K in nv/v 500-79 525 1000 24 1049-500 30 525-1000 -75 1049
1200 CW direction, TN 1000 N m Deviation from KRISS values in nv/v 900 600 300 0-300 -600-900 -1200 KRISS 500 N m NIS 500 N m Lab Results KRISS 1000 N m NIS 1000 N m Figure 2: Deviations of measured values on CW direction from the pilot. 1200 CCW direction, TN 1000 N m Deviation from KRISS values in nv/v 900 600 300 0-300 -600-900 -1200 KRISS -500 N m NIS -500 N m Lab Results KRISS -1000 N m NIS -1000 N m Figure 3: Deviations of measured values on CCW direction from the pilot.
2.6.2 Linking key comparison CCM.T-K1.3 to key comparison CCM.T-K1 It is well-known that there is no absolute reference value to circulate in the field of torque metrology since the sensitivity of the torque transducers varies over time. Thus, only relative deviations can be compared. In addition, for reporting the results and calculating degree of equivalence, the deviations and their uncertainties were converted into torque units using the equation (4) and (5) and are given in Table 6, N-K = N-K KRI (4) ( N-K ) = ( N-K ) KRI (5) Table 6: Deviations of the measured values of NIS from the pilot, KRISS Torque Deviation from the pilot N-K in mn m N-K in mn m 500-45.0 299.3 1000 13.4 597.4-500 -16.9 299.2-1000 42.9 597.5 Now, the degree of equivalence of NIS, N to the KCRV can be calculated as follows: = N-K K-KCRV (6) ( )= ( N-K )+ ( K-KCRV ) (7) In order to link the results of CCM.T-K1.3 to those of CCM.T-K1, the degree of equivalence of the link institute, KRISS to the key comparison reference value (KCRV), D K-KCRV should be calculated first. Since the degree of equivalences were reported for each of two transfer artifacts separately [1], it is necessary to combine the results from KRISS at each torque. It seems reasonable that this should be done by taking a weighted mean of KRISS results, giving more weight to the values with lower uncertainties. The uncertainties of the weighted means should be evaluated by considering correlation
between results from two transfer artifacts because two measurements had been made using the same torque standard machine. The weighted means and their corresponding uncertainties were calculated according to equation (8) and (9) [4]., = K,TB2 K,TB2 + K,TT1 K,TT1 K, TB2 + K.TT1 (8) where, w K,i = 1/u 2 (x K,i ) and i= TB2 and TT1 (, )= ( K,TB2 ) ( K,TT1 ) ( K,TB2 ) ( K,TT1 ) K,TB2 + K,TT1 +2 ( K,TB2, K,TT1 ) ( K,TB2 ) ( K,TT1 ) (9) The correlation coefficient can be calculated by the following equation: K,TB2, K,TT1 = ( STD ) K, TB2 K, TT1 (10) where, ( STD ) is the standard of the KRISS torque standard machine. The original KRISS results and combined results are listed in Table 7. Table 7: KRISS results of two transfer artifacts on CCM.T-K1 [1] and overall results. The relative expression of uncertainties in [1] is converted in torque units of N m for the calculation of degree of equivalences. TB2 TT1 Overall Torque K ( K ) K ( K ) K ( K ) 500 500.015 0.018 499.977 0.033 500.007 0.018 1000 1000.036 0.034 999.947 0.073 1000.020 0.034-500 -500.014 0.018-499.970 0.032-500.003 0.018-1000 -1000.036 0.035-999.941 0.069-1000.017 0.036
The KCRVs of CCM.T-K1 were also combined for linking and calculated as listed in Table 8. The uncertainties were calculated assuming there is no correlation between two KCRVs. Table 8: KCRVs of two transfer artifacts on CCM.T-K1 [1] and overall KCRVs Torque ref TB2 TT1 Overall ( ref ) in mn m ref ( ref ) in mn m ref ( ref ) in mn m 500 500.000 2.9 500.000 3.9 500.000 2.3 1000 1000.000 5.5 1000.000 8.1 1000.000 4.6-500 -500.000 3.0-500.000 3.6-500.000 2.3-1000 -1000.000 5.8-1000.000 7.5-1000.000 4.6 The overall degree of equivalence of KRISS in CCM.T-K1 can be computed from the relations K-KCRV = K ref (11) ( K-KCRV )= 2 ( K ) ( ref ) (12) The degree of equivalence of NIS was computed using equations (6), (7), (11) and (12) and the results are listed in Table 9 along with the overall degree of equivalence of KRISS in CCM.T-K1. The figures 4 and 5 graphically show the results. Table 9: The overall degree of equivalence (DOE) of KRISS in CCM.T-K1 [1] and degree of equivalence of NIS Torque Overall DOE of KRISS K-KCRV in mn m ( K-KCRV ) in mn m in mn m NIS DOE ( ) in mn m 500 6.9 17.2-51.8 299.7 1000 20.1 33.6-6.7 598.3-500 -3.4 17.5-13.5 299.7-1000 -16.6 34.4 59.5 598.5
600 CW direction, TN 1000 N m 450 Degree of Equivalence in mn m 300 150 0-150 -300-450 -600 KRISS 500 N m NIS 500 N m Lab Results KRISS 1000 N m NIS 1000 N m Figure 4: The degree of equivalence of KRISS and NIS results in CW direction to the KCRVs. 600 CCW direction, TN 1000 N m 450 Degree of Equivalence in mn m 300 150 0-150 -300-450 -600 KRISS -500 N m NIS -500 N m Lab Results KRISS -1000 N m NIS -1000 N m Figure 5: The degree of equivalence of KRISS and NIS results in CCW direction to the KCRVs.
2.7 Summary Bilateral comparisons between KRISS and NIS were conducted at the torques of 0 N m, 500 N m and 1000 N m as an extension of Key Comparison CCM.T-K1. Analysis reveals that all results of NIS are equivalent to the KCRVs within their uncertainties. As clearly shown in Figure 2 and 3, all NIS results are matched with those of KRISS within their claimed uncertainties. 3. References [1] Dirk Röske, Final report on the torque key comparison CCM.T-K1. Measurand torque: 0 N m, 500 N m, 1000 N m, Metrologia 46 (2009), Tech. Suppl., 07002. [2] M. G: Cox, The evaluation of key comparison data, Metrologia 39 (2002), 589-595. [3] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP and OIML, 1995 Guide to the Expression of Uncertainty in Measurement, 2 nd edn (Geneva: International Standards Organization) [4] Kazunaga Ueda, Thomas W Bartel, Goh Hok Liok, Lee Hsin-Tse, John Man, Yon- Kyu Park and Seung-yin Wong, Final report on APMP.M.F-K4.b key comparison for 2 MN force, Metrologia 49 (2012), Tech. Suppl., 07008.
Appendix: Comparison on measurand torque at 0 N m, 20 N m, 50 N m, 1 kn m and 2 kn m This part describes the results of auxiliary torque comparisons in the range of 0 N m, 20 N m, 50 N m, 1000 N m and 2000 N m, which was conducted in parallel with CCM.T-K1.3. These results are not a part of the CCM.T-K1.3 comparison, but included in this report in order to show the equivalence between KRISS and NIS at the torque range not covered by CCM.T-K1.3. These auxiliary comparisons were conducted using the same protocol as that of CCM.T-K1.3. A1.Torque standards KRISS used the same torque standard machine as used in CCM.T-K1.3, but NIS used a different, vertical-type torque standard machine with 3 kn m capacity. The TSMs used in these comparisons are listed in Table A1. Table A1: Torque standard machines used in the torque comparison Torque Standard Machine Institute Capacity Note Type of / kn m applied torque / % KRISS 2 Deadweight 0.005 Pilot Lab. NIS 3 Comparator (Vertical) 0.05 Reference transducers traceable to KRISS A2.Traveling standards Torque comparisons were conducted at 0 N m, 20 N m, 50 N m, 1 kn m and 2 kn m using two different travelling standards with a capacity of 50 N m and 2 kn m, respectively, which belong to NIS. On the other hand, the same bridge amplifier (DMP40) was used as CCM.T-K1.3.
Table A2: Travelling standards used in the torque comparisons at 0 N m, 20 N m, 50 N m, 1 kn m, and 2 kn m Capacity Type Maker Serial # Adaptation Comparison points 50 TN HBM 014840008 50 h7 80 shaft 20, 50 2000 DmTN GTM 52331 50 h7 80 shaft 1000, 2000 A3. Comparison results The reported results are listed in Table A3 and A4, respectively. Table A3: Reported deflections and relative uncertainties (k = 2) at 20 N m and 50 N m torques Institute Date Deflection Clockwise torque 20 N m 50 N m Deflection KRISS 2010-09-16 0.609096 0.005 1.522834 0.005 NIS 2010-12-14 0.609093 0.05 1.522825 0.05 Counterclockwise torque 20 N m 50 N m Institute Date Deflection Deflection KRISS 2010-09-17-0.609135 0.005-1.522879 0.005 NIS 2010-12-15-0.609072 0.05-1.522788 0.05
Table A4: Reported deflections and relative uncertainties (k = 2) at 1 kn m and 2 kn m torques Institute Date Deflection Clockwise torque 1 kn m 2 kn m Deflection KRISS 2010-11-05 1.001399 0.005 2.003316 0.0051 NIS 2010-12-14 1.001329 0.05 2.003278 0.05 Counterclockwise torque 1kN m 2kN m Institute Date Deflection Deflection KRISS 2010-11-08-1.001658 0.0051-2.004343 0.0052 NIS 2010-12-15-1.001528 0.05-2.004563 0.05 Figure A1-A2 and A3-A4 show the relative deviations of NIS results with respect to KRISS results for 50 N m and 2000 N m transfer artifacts, respectively. As clearly shown in Figure A1-A2 and A3-A4, NIS results are matched with those of KRISS within the claimed uncertainties. Furthermore, the relative uncertainties of KRISS are almost ten times better than those of NIS. Thus, the weighted means and their uncertainties are almost same as KRISS means and uncertainties. Thus, it would be enough to decide the equivalence of the results by simply comparing NIS results with respect to KRISS results.
deviation from KRISS values / 10-5 60 40 20 0-20 -40-60 CW direction, TN 50 N m KRISS 20 N m NIS 20 N m KRISS 50 N m NIS 50 N m Lab Results Figure A1: Comparison results at 20 N m and 50 N m in CW direction. 60 CCW direction, TN 50 N m deviation from KRISS values / 10-5 40 20 0-20 -40-60 KRISS 20 N m NIS 20 N m KRISS 50 N m NIS 50 N m Lab Results Figure A2: Comparison results at 20 N m and 50 N m in CCW direction.
60 CW direction, DmTN 2 kn m deviation from KRISS values / 10-5 40 20 0-20 -40-60 KRISS 1 kn m NIS 1 kn m KRISS 2 kn m NIS 2 kn m Lab Results Figure A3: Comparison results at 1 kn m and 2 kn m in CW direction. 60 CCW direction, DmTN 2 kn m deviation from KRISS values / 10-5 40 20 0-20 -40-60 KRISS 1 kn m NIS 1 kn m KRISS 2 kn m NIS 2 kn m Lab Results Figure A4: Comparison results at 1 kn m and 2 kn m in CCW direction.