Four half-wave resonant, air-backed, x-cut quartz disc transducers designed and fabricated at the National Bureau of Standards (NBS) were circulated among seven laboratories between February 1978 and June 1980. Participants were given detailed instructions, including values and tolerances for the amplitude and frequency of the sinusoidal voltages to be applied to each transducer. Five distinct methods were used to measure the total continuous-wave beam power radiated by each transducer into a reflectionless water load. Two independent electrical techniques were used to directly measure radiation conductance. A detailed tabulation of the results was published by the NBS in 1983. Of the 1160 measurements of power or radiation conductance constituting the result of this collective effort, 90% were made using four techniques based on radiation pressure. In this paper, current methods of uncertainty analysis are applied to a representative sample of data from radiation-pressure measurements. Old and new analyses are compared to assess the sensitivity of results to the choice of computational methods. Newly computed combined standard uncertainties are smaller than the original uncertainties for more than 60% of the data considered. For the remaining data, the newly computed uncertainties are considerably larger than the original values; sufficiently large to make the average effect an increase of more than 15 %.
Pub Type: Journals
international comparison, ultrasonic measurements, Ultrasonics