Final Report on APMP.M.M-K4.1 - Bilateral Comparison of 1 kg Stainless Steel Mass Standards between KRISS and A*STAR Jin Wan Chung 1, Shih Mean Lee, Sungjun Lee 1 1 Korea Research Institute of Standards and Science, Daejeon 34113, Rep. of Korea National Metrology Centre (NMC), Agency for Science, Technology and Research (A*STAR), 1 Science Park Drive, Singapore 1181, Singapore Abstract: This report describes a bilateral comparison of 1 kg stainless steel mass standard between the Korea Research Institute of Standards and Science (KRISS), and the National Measurement Centre, Agency for Science, Technology and Research (A*STAR), Singapore, which is registered in the BIPM KCDB as APMP.M.M-K4.1. This nominal value was chosen as it followed the nominal value of CCM.M-K4. This bilateral comparison was piloted by KRISS. The transfer weight was delivered to A*STAR from KRISS in April 015 and received back from A*STAR in October 015. A*STAR have the aim to improve their calibration measurement capability (CMC) for mass of 1 kg. KRISS participated in both this comparison and CCM.M-K4, providing a link between the A*STAR result and the KCRV of CCM.M-K4. The final report of this comparison is being given in this document. Final Report on APMP.M.M-K4.1 1 of 11
1. Introduction This report describes a bilateral comparison of a 1 kg stainless steel mass standard between KRISS (Korea), and A*STAR (Singapore). The objectives of the comparison were to facilitate the demonstration of metrological equivalence between the two participating national laboratories, and to check the capability of A*STAR to improve quoted calibration measurement capability (CMC).. Transfer standards The transfer standard for this comparison comprises 1 kg stainless steel cylindrical weight with a knob (OIML R111 design []). The weight was housed in a wooden container. It was produced by Mettler-Toledo Co. of Switzerland and purchased by KRISS in 00 as shown Figure 1. The physical properties such as volume and density of the transfer standard and associated uncertainty are given by KRISS and shown in the Table 1. The magnetic susceptibility of the transfer standard was not measured. Figure 1. Travelling container and its contents for transfer standard Table 1. Physical properties of standard Identification of Standard Density (0 ), kg m -3 Volumetric Uncertainty of density (k = ), kg m -3 coefficient of Expansion, K -1 8069 8011.095 0.01 0,000 050* *Assumed Final Report on APMP.M.M-K4.1 of 11
3. Participants and Contributors Participant 1 (Pilot laboratory) Korea Research Institute of Standards and Science (KRISS), 67 Gajeong, Yuseong, Daejeon 34113, Rep. of Korea Phone: +8-4-868-5111 Fax: +8-4-868-501 Contact person: Jin Wan Chung (Email: jwchung@kriss.re.kr) Participant National Metrology Centre (NMC), Agency for Science, Technology and Research (A*STAR), 1 Science Park Drive, Singapore 1181, Singapore Phone: +65 679 1960 Fax: +65 679 199 Contact person: Shih Mean Lee (Email: lee_shih_mean@nmc.a-star.edu.sg) Documentation & Supports Korea Research Institute of Standards and Science (KRISS), 67 Gajeong, Yuseong, Daejeon 34113, Rep. of Korea Phone: +8-4-868-5815 Fax: +8-4-868-501 Corresponding: Sungjun Lee (Email: lsjun@kriss.re.kr) 4. Circulation scheme and time schedule The transfer standard was hand-carried by KRISS personnel from KRISS to A*STAR and A*STAR personnel from A*STAR to KRISS respectively with ATA carnet. No special custom procedures were required between Korea and Singapore. The agreed and realized time schedule was as follows: - Measurements at KRISS, Daejeon: April 015 (marked as KRISS_1 in Result section) Transport to A*STAR, Singapore - Measurements at A*STAR, Singapore: May to June 015 (A*STAR) Transport to KRISS, Daejeon Final Report on APMP.M.M-K4.1 3 of 11
- Measurements at KRISS, Daejeon: Nov. 015 (marked as KRISS_ in Result section) 5. Technical protocol The agreed technical protocol for APMP.M.M-K4.1 is the same as that of CCM.M-K1. 6. Results As shown in the technical protocol, a completed measurement report containing the measurement result, data of the ambient conditions, list of instrument used and a description of the traceability of the participant's reference standard was submitted to KRISS, which was acting as a collector of all result in the comparison. After all measurements were completed, the measurement reports were sent to KRISS, which provided the link to CCM.M-K4 through APMP.M.M-K4.1. 6.1 Measurement Results Table and Figure show the results of the reported mass value and their combined standard uncertainties as given by A*STAR and KRISS. The value m A (= m Areported - m 0 ) is the deviation from the nominal mass (m 0 ) value and u c is combined standard uncertainty (k=1) claimed by each laboratory. The number of significant digits has been restricted to a maximum of three for m A and two for u c [3]. As is normal practice with mass key comparisons, all mass values are true mass(and not conventional mass) Table. Measurement results for transfer standards reported from A*STAR and KRISS Labs. Date m A / mg 1 k g u c / mg KRISS _1 Apr. 015 0.004 0.017 A*STAR May 015-0.04 0.05 KRISS _ Nov. 015-0.035 0.017 Final Report on APMP.M.M-K4.1 4 of 11
Figure. Measurement result for transfer standard The stability of the transfer standard 1 kg artifact was assessed by KRISS by measuring the mass values of the transfer standard and a check standard as against one of KRISS s reference standards before and after the measurement loop. The results are shown in Table 3. As shown Table 3, check standard of KRISS shows no drift, but the transfer standard shows drift of - 0.039 mg. This was taken into account when a KCRV for this comparison was determined. Averaged mass value between KRISS_1 and KRISS_ is used as the KCRV in this comparison and instability uncertainty of ± 0.011 mg (see equation (1)) is included. Table 3. Mass change of transfer standard and check standard as measured by KRISS during the comparison Nominal Mass Mass change of check standard between Apr. and Nov., 015 / mg Transfer standard KRISS Check Standard 1 kg -0.039-0.001 Final Report on APMP.M.M-K4.1 5 of 11
From the measurement, instability of transfer standard was found as shown Table 3. The uncertainty (k=1) due to instability of transfer standard (u inst. ) during Apr. to Nov., 015 was calculated using equation (1) as a rectangular distribution and obtained as 0.011 mg: u inst. ( m KRISS _ m KRISS _1 ) = (1) ( 3) Where: m _KRISS1 : Mass measurement by KRISS in Apr. 015 m _KRISS : Mass measurement by KRISS in Nov. 015 The corrected measurement results for transfer standard reported from A*STAR and KRISS are shown in Table 4. Table 4. Calculated KCRV and uncertainty and the measurement results reported by A*STAR for the transfer standards Labs. KCRV (Averaged KRISS_1 and KRISS_) m A / mg 1 k g u c / mg -0.016 0.00 A*STAR -0.04 0.05 KRISS has thus provided the KCRV for APMP.M.M-K4.1 and therefore have a zero difference from this value for their averaged result. The difference between this KCRV and the result reported by A*STAR is then: (-0.016) mg - (-0.04) mg = +0.008 mg. 7. Linkage to CCM.M-K4 The analysis follows the method used in the APMP.M.M-K1 [4],[5] for linkage to the key comparison reference value (KCRV) obtained in the CCM 1 kg key comparison CCM.M-K4. Final Report on APMP.M.M-K4.1 6 of 11
The aim is the analysis is to estimate the degree of equivalence for participating laboratory; NMC, A*STAR through the link laboratory KRISS. 7.1. Linked Results The measurement data were analyzed using the method in [4], [5]. This method links APMP.M.M-K4.1 to CCM.M-K4 and gives the degrees of equivalence required by the MRA [1] without the need for an APMP.M.M-K4.1 reference value. The result of analysis was checked doubly for each mass standard. From the final report on CIPM key comparison of 1 kg standards in stainless steel (CCM.M- K4) [6]: KRISS-KCRV = -0.01 (0.09) mg () Where: The number in the parenthesis is the expanded uncertainty with a coverage factor k=. This result is interpreted as the mass value assigned by KRISS to the mass of the 1 kg transfer standard is 0.01 mg less than the KCRV. This difference from KCRV was determined in the course of the key comparison CCM.M-K4. The KCRV for CCM.M-K4 was estimated using the generalized linear least-squares estimation (GLS) value of the results of participating laboratories for the mass of 1 kg transfer standards. The equation describing each comparison measurement for transfer standard can be written as: m A( Labi ) p m0 = i ( m0 m A) + ei, p (3) Where: m A (Lab i ) p : p th value assigned to transfer standard by laboratory i m 0 : nominal mass of transfer standard m A : Mass of transfer standard Δ i : p th value assigned to transfer standard by laboratory i m A (Lab i ) p : Bias of laboratory e i,p : random error associated with the measurement Final Report on APMP.M.M-K4.1 7 of 11
Similarly, the equation for the result of the link laboratory is: m Lab ) K = ( K m ) + e cx ( (4) i i cx i Where: m cx (Lab i )-K: Measured deviation between the link laboratory i and the KCRV at m 0 m cx : GLS mass values of the CCM.M-K4 transfer standards m A : Mass of transfer standard K: Key comparison reference value e i : random error associated with the measurement The known values are m x (Lab i ) p - m 0 and m c (Lab i )-K. To solve these equations, a constraint is required and we choose K -m c =0; so that the value obtained for Δ i from the solution is the expected deviation of the laboratory s result from the KCRV. New correction value for national prototype kilogram (No. 7) given by BIPM is adopted to KRISS reference standard. The results are shown in Table 5 and Figure 3. Table 5 and Figure 3 give the calculated deviations from the KCRV of CCM.M-K4 for both laboratories, together with the associated uncertainties. Table 6 gives the difference in assigned mass values between KRISS and A*STAR, NMC and its uncertainty at 1 kg. Table 5. Deviation from the KCRV (Key Comparison Reference Value of CCM.M-K4) Labs. Difference from (CCM) Expanded uncertainty KCRV / mg (k=),/ mg KRISS -0.01 0.034 A*STAR -0.00 0.050 En no. 0.13* * See equation (6) Final Report on APMP.M.M-K4.1 8 of 11
Figure 3. Mass values assigned by KRISS and A*STAR, NMC with bars representing the expanded uncertainties (Zero value corresponds to the KCRV of CCM.M-K4) Table 6 Difference in assigned mass value between KRISS and A*STAR, NMC and the expanded uncertainty for 1 kg in mg NMIs Difference in assigned mass / mg Expanded uncertainty (k=), U diff / mg A*STAR-KRISS -0.008 0.041 As shown in Table 6, the expanded uncertainty for mass difference between KRISS and A*STAR is obtained by using equation (5); U diff u = ( u = diff A* STAR + u inst. u KRISS (5) ) For check equivalence between both NMIs, En number is calculated by using equation (6) with the results given in Table 5; Final Report on APMP.M.M-K4.1 9 of 11
E n A B = (6) U + U KRISS A* STAR Where: A is the KRISS value from this comparison, B is the A*STAR value from this comparison U KRISS is the expanded uncertainty of KRISS with coverage factor k=, U A*STAR is the expanded uncertainty of A*STAR with coverage factor k=. As a result, En number is 0.1. That means that both NMI s mass value for 1 kg has a good degree of equivalence within the given uncertainty. 8. Conclusion The results of KRISS and A*STAR/NMC are consistent at the nominal mass value of 1 kg. The A*STAR result is also consistent with the KCRV of CCM.M-K4 with k =. This can be seen in Fig. 3 where the expanded uncertainty bar for the result cross each other and the line of zero mass (i.e. the KCRV). Acknowledgments The author wishes to thank Dr. Stuart Davidson, NPL for useful discussions and comments on the manuscript and Mr. Kwang Pyo KIM, KRISS for support and preparation on measurement. Final Report on APMP.M.M-K4.1 10 of 11
References [1] Mutual recognition of national measurement standards and of calibration and measurement certificates issued by national metrology institutes, BIPM, Paris, 14 October 1999. [] Weights of classes E 1, E, F 1, F, M 1, M, M 3, International recommendation OIML R 111, OIML, Paris 004. [3] Guide to the Expression of Uncertainty in Measurement, International Organization for Standardization, Geneva, Switzerland, 1993. [4] Veera et al.: Final report on APMP comparison of 1 kg mass standards (APMP.M.M- K1), Metrologia, Tech, Suppl., 41, 07003 (004). [5] C.M. Sutton, Analysis and linking of international measurement comparisons, Metrologia 41, 7-77 (004). [6] Luis Omar Becerra et al.: Final report on CIPM key comparison of 1 kg stainless steel mass standards (CCM.M-K4), Metrologia, Tech, Suppl., 51, 07009 (014). Final Report on APMP.M.M-K4.1 11 of 11