Mark Ballico

Limitations of the Welch-Satterthwaite approximation for measurement uncertainty calculations

The Welch-Satterthwaite approximation is used to estimate an effective degrees of freedom for a probability distribution formed from several independent normal distributions for which only estimates of the variance are known. The International Organization for Standardization (ISO) Guide to the Expression of Uncertainty in Measurement recommends its use for the calculation of expanded uncertainties (the ISO term for confidence intervals) for uncertainties formed from several distributions. Although it is well recognized that this formula is an approximation, counter-intuitive results may be obtained when a variance estimate of the dominant distribution has a small number of degrees of freedom; the calculated confidence limits are sometimes found to decrease with increases in the contributing uncertainties.

Uncertainty Budgets for Realization of ITS‐90 by Radiation Thermometry

Recent international comparisons [1,2] and key comparisons have shown that the realization of the International Temperature Scale of 1990 (ITS‐90) above the freezing point of silver and its dissemination is more difficult than expected. In many cases, the deviations of the local scale realizations were larger than the combined estimated uncertainties could reasonably justify. On the other hand, it must be considered that the realization of the ITS‐90 by radiation thermometry is a complex exercise involving a large number of operations with many influencing parameters. Furthermore, the key comparisons need a unified approach to the treatment of uncertainties. Consequently, a rigorous standard approach for the calculation of uncertainties is necessary. In this paper three different operational schemes have been identified for realizing the ITS‐90 by radiation thermometry. For all three schemes an analysis is presented of the baseline parameters underlying the scale realization above the freezing point of si...

Modelling of the effective emissivity of a graphite black body

Measurements of the effective radiance temperature of a high-temperature graphite tube black-body furnace are compared with those calculated from two theoretical models. The first model extends the conventional annular zone model by including effects of vignetting and temperature gradients. Because the black-body surface is not a perfect diffuse reflector, this is found to be insufficiently accurate and a Monte-Carlo integration with a diffuse/Fresnel-specular surface model is presented, which gives good agreement with the measured data.

Novel Experimental Technique for Measuring High-Temperature Spectral Emissivities:

A novel experimental technique for the measurement of spectral emissivities in the temperature range 500 to 1000°C of materials that are both poorly thermally conducting and have a high level of transparency is presented. A Fourier transform infrared spectrometer (FT-IR) is used to compare the spectral radiance from a nearly isothermal, but rapidly cooling, sample to that of a reference blackbody. Experimental results obtained from metal metaborate samples and a theoretical analysis of the technique show the effectiveness of this method, particularly for the case of molten samples.

A Study of the Temperature Dependence of Inhomogeneity in Platinum‐Based Thermocouples

It is well known that contamination of platinum‐based thermocouples due to use at high temperatures causes the local Seebeck coefficient of the wire to change from its ‘initial’ state. As this exposure and contamination is usually not uniform along the length of the thermocouple, the Seebeck coefficient also becomes a function of position along the thermocouple, leading an exposure‐dependent thermoelectric signature or inhomogeneity. At NML (CSIRO) we have, for many years, included explicit measurement of the thermocouple inhomogeneity in all calibrations of ‘used’ thermocouples, using a special uniform‐temperature ‘scan furnace’ having an entrance region with a steep gradient. The thermocouple inhomogeneity value is used in the calculation of the calibration uncertainty: indeed it is usually the dominant component, as the uncertainty in the reference standard SPRTs and fixed points is usually negligible by comparison. The inhomogeneity measurements at NML are usually done at 450 °C, low enough to limit d...

The ‘Mini‐Coil’ Method for Calibration of Thermocouples at the Palladium Point

The melting point of palladium (1554.8 °C on the ITS‐90) has widespread use as a reference point for the calibration of thermocouples to 1550 °C. The ‘wire bridge’ method is widely used to realize this reference temperature. At NML (CSIRO) a more convenient technique, the ‘mini‐coil’ method, in which a small coil of Pd wire is squeezed over the thermocouple junction has been developed and used for some years. In the study presented here the ‘mini‐coil’ method is assessed for any possible systematic errors, such as arising from the argon gas purge or the heating rate, and is compared with the ‘wire bridge’ method. The two methods are found to give equivalent results; however the larger mass of the Pd used in the mini‐coil method gives longer, better defined and more reproducible melting plateaus (reproducibility is 0.5 μV). Analysis of possible Pt‐Pd inter‐diffusion is examined by comparing melting plateaus obtained after heating the thermocouple and Pd sample for various periods just below the Pd melting ...

Final report on supplementary comparison APMP-T-S3-03 of industrial platinum resistance and liquid in glass thermometers from -40 °C to 250 °C

Industrial thermometers such as industrial platinum resistance thermometers (iprts) and liquid-in-glass thermometers (LIGTs) are widely used in industry. Because Key Comparisons are limited to direct realizations of ITS-90, and not all APMP NMIs have participated in them, the national metrology institutes (NMIs) of Thailand and Australia (NIMT and NIMA) organized an APMP supplementary comparison to support the approval of CMCs (calibration and measurement capabilities) for these laboratories. The comparison, performed in 2003, covered the range from -40.0 °C to 250.0 °C, using IPRTs (Hart Scientific 5626-12-S), total immersion (ASTM 62C, 120C) and partial immersion (ASTM 40C) LIGTs. Ten NMIs from the APMP: KIM-LIPI (Indonesia), ITDI (Philippines), MSL (New Zealand), NBSM (Nepal), NMIA (Australia), NIMT (Thailand), SCL (Hong Kong), SIRIM (Malaysia), SPRING (Singapore) and VMI (Vietnam) were divided into two loops to shorten the circulation time, and these were linked by the two pilot laboratories. This report describes details of the artifacts, the circulation schedule, the measurement procedures, the results submitted by participants, uncertainties and the analysis of the results. Reference values calculated using simple mean, median and weighted mean were consistent with each other, but as the Birge criterion was satisfied, the weighted mean with its lower uncertainty was adopted. The artifacts were found to be stable over the comparison and the results of the loop linking labs consistent, allowing an uncertainty of 2 mK to 4 mK to be achieved for the IPRT reference value and 10 mK to 20 mK for the LIGT reference values. These uncertainties allowed the comparison data to be used to adequately test the uncertainties of all the participant laboratories, and hence to directly support their CMC claims. Main text. To reach the main text of this paper, click on Final Report. Note that this text is that which appears in Appendix B of the BIPM key comparison database kcdb.bipm.org/. The final report has been peer-reviewed and approved for publication by the APMP, according to the provisions of the CIPM Mutual Recognition Arrangement (MRA).

A Simple Experimental Technique for Estimation of the Band‐pass of Infrared Radiation Thermometers

The infrared radiation thermometer is now becoming a very widely used instrument for industrial temperature measurement and hazardous temperature assessment. Its simple design allows a very low manufacturing cost, and hence its widespread adoption as a convenient non‐contact thermometer. The spectral sensitivity of these instruments is usually set by the transmission properties of the windows on the pyrometer, and, although the manufacturer usually makes some statement as to the operating wavelength region, it is difficult to know exactly what is being reported. Calibration laboratories often require knowledge of the band‐pass for appropriate calculations of uncertainty, corrections for blackbody emissivity, or, as is commonly becoming required, correction of blackbody radiance temperatures for the calibration of cheaper ‘fixed emissivity (eg. 95%)’ pyrometers. An explicit measurement of the spectral response of the pyrometer would require a tunable IR source, and special attention to the effects of ambie...