Reference Viscosity of Argon at Low Density in the Temperature Range from 290 K to 680 K

Abstract

An all-quartz oscillating-disk viscometer of very high precision was used to measure the temperature dependence of the viscosity of argon in the limit of zero density. The measurements were based on a single calibration value at 298.15 K, which was calculated theoretically using an accurate ab initio pair potential for argon and the kinetic theory of dilute gases. The uncertainty of the experimental data is conservatively estimated to be 0.15 % at room temperature increasing to 0.2 % at the highest temperature of 680 K. The new data, as well as viscosities determined in 2007 at the National Institute of Standards and Technology in the range from 200 K to 400 K, agree excellently within ±0.1 % with the theoretically calculated viscosity of argon at zero density. On the contrary, the widely accepted viscosity data recommended by Kestin et al. (J Chem Phys 56:4119, 1972) deviate by as much as 0.9 %.

Towards Reference Viscosities of Carbon Monoxide and Nitrogen at Low Density Using Measurements between 290K and 680K as well as Theoretically Calculated Viscosities

“…In addition, a comparison of our experimental argon data between 291 K and 682 K [8] and of experimental reference values obtained by May et al [19] at NIST (203 K to 394 K) with the theoretical argon values [12] showed excellent agreement within ±0.1 %. Considering further the consistency of the temperature functions of the theoretical viscosity values for the helium and argon pair potentials, the performance of our all-quartz oscillating-disk viscometer is well established.…”

“…The principal difference between the old reference values for helium used for the calibration in our earlier measurements on nitrogen in 1984 [20,21] [16]. According to the sound error analysis in References [8,15], the relative uncertainty of the measurements with our all-quartz oscillating-disk viscometer is conservatively estimated to be ±0.15 % at room temperature and up to ±0.20 % at higher temperatures assuming that the uncertainty could be slightly increased with temperature. The total uncertainty in the viscosity, which additionally includes the allocation errors of the temperature and density measurements considering the temperature and density dependencies of the viscosity, corresponds practically to the uncertainty of the viscosity measurements alone, as the contributions of the allocation errors are negligibly small.…”

“…For our high-precision measurements on argon at low density between 290 K and 680 K [8], we adopted as a reference value a theoretically calculated zero-density viscosity coefficient η (0) at 298.15 K for argon together with its temperature derivative near room temperature [12], resulting from the kinetic theory of dilute gases and the new ab initio argon-argon interatomic potential [13,14]. For the subsequent measurements on methane and hydrogen sulfide [15] in the same temperature range, we changed the calibration in so far that we applied a theoretical viscosity value for helium [16] based on a highly accurate ab initio helium pair potential [17].…”

“…The values used for the calibration follow from the zero-density viscosity coefficient at 298.15 K, the temperature derivative near room temperature and the density derivative for low densities (see References [8,15]). The principal difference between the old reference values for helium used for the calibration in our earlier measurements on nitrogen in 1984 [20,21] [16].…”

“…The drawing and the characteristics of the newly constructed viscometer were specified in Fig. 1 and Table 1 of Reference [8]. Principles of its construction were given in an earlier article [9], whereas details of the manufacturing were changed over the years to facilitate the assembly of the instrument.…”