No abstract

The absolute Seebeck coefficient of platinum was determined from 80 to 340 K by direct comparsion to lead. Results of this comparison disagree with previous results which have been used for the calculation of absolute values for other materials. The thermal conductivity λ and electrical resistivity ρ of the lead standard were also determined. The electrical resistivity could be described with a modified Gruneisen-Bloch equation which allows for the effect of thermal expansion on the Debye temperature ΘD. The ratio λρ/T was within 1% of the Sommerfeld value of 2.443×10−8 (V/K)2 from 1.0 to 5.0 ΘD.

The effects of temperature, purity, magnetic state, and crystal structure on the thermal conductivity, electrical resistivity, and Seebeck coefficient of iron were obtained from measurements on Armco iron (99.5% pure, ρ300/ρ4.2=11.0) and a high-purity iron (99.95% pure, ρ300/ρ4.2=26.2). The most probable determinate errors of the measurements were thermal conductivity ±1.5%, electrical resistivity ±0.1%, and Seebeck coefficient ±0.9%; and larger absolute errors. Where theory permits, the thermophysical properties of iron are discussed in terms of contributing transport mechanisms. The thermal conductivity of iron can be calculated to ±1.5% between 0° and 910°C from electrical-resistivity measurements and the lattice portion of the thermal conductivity determined in this study.

The thermal conductivities, h, of single-crystal and polycrystalline UO, were measured from 80" to 420°K. The results indicate no observable difference in h between single-crystal and polycrystalline UO,, and both materials have broad peaks in x a t ~2 2 0°K . The results were used with literature values to determine the effect of closed porosity on A. The thermal conductivity of theoretically dense UO, is described phenomenologically from 80" to 140O0K, where conduction is dominated by the phonon component. The phonon conduction is analyzed by comparison with Tho,. This analysis indicates that the high-temperature x is limited by 3-phonon Umklapp scattering processes. Scattering by the disordered spins associated with the paramagnetic U ions contributes a large temperatureindependent phonon scattering term. This mechanism has a mean free path of about 51 A, which implies that grain boundaries and impurities have a relatively insignificant effect on the phonon conduction far above the antiferromagnetic-paramagnetic transition a t -30°K. This implication agrees with the experimental results.

The thermal conductivity λ, electrical resistivity ρ, and Seebeck coefficient S of a high-purity Cr specimen (ρ273/ρ4.2=380) were measured from 285 to 1000 K. The ρ and S of two other Cr specimens (ρ273/ρ4.2=380 and 58) were determined from 300 to 1300 K. The ρ and S results from the three specimens are in excellent agreement and all three properties agree to within experimental uncertainty with previous low-temperature results on the same specimens over the temperature range of overlap. Near T′N (300–320 K), the present λ results are within 0.7% of the previous data and indicate that λρ/T should be smooth to within 1%. At high temperature, the present λ data are about 8% above those of Powell and Tye but the ratios of λρ/T agree to within 2% up to 1000 K. These new data on pure Cr are compared to calculations from standard transport theory and to previous results from W and Mo.

Thermal conductivity λ, electrical resistivity ρ, and Seebeck coefficient S measurements have been made from 80 to 400 K on a Ni3Fe alloy (74.77-at. % Ni and 25.33-at. % Fe) in both disordered and highly ordered states. The effect of lattice disorder is to lower λ and S and increase ρ. Although λ and ρ for the three states studied differ by about 50%, the Lorenz ratios λρ/T are the same to within the combined uncertainty of the measurements. This ratio is near the Sommerfeld value Lo from 200 to 400 K, and the positive deviation from Lo below 200 K indicates a significant lattice component to λ in all states of disorder. The thermal conductivities of the three states of Ni3Fe exhibit peaks between 142 and 175 K. The unusual relative shift of the temperatures of the maximum conductivities is related directly to the large deviation of the electrical resistivity differences from Matthiessen's rule.

A radial heat flow technique was used to measure the thermal conductivity, k, of polycrystalline UO2in the range ‐57° to 1100°C. The technique yielded results with a probable accuracy of ±1.5% and a precision of ±0.1% in the range 50° to 1100°C. Meaningful measurements were limited to 1100°C by Pt‐90 Pt10Rh thermocouple instability, although the apparatus was structurally sound to 1400°C. The thermal conductivity data up to 1000°C could be explained on the basis of heat transport by phonons. The thermal resistance, l/k, exhibits a linear temperature dependence from 200° to 100°C, which is expected for an insulator well above the Debye temperature. The slope of the l/k‐temperature plot is 0.0223 cm w−1 which is independent of impurity content and is associated with three phonon umklapp processes. The intercept is sensitive to impurity content as indicated by the fact that it was decreased by a decrease in the oxygen/uranium ratio. Between 1000° and 1100°C, there is a slight departure of l/k from linearity which may be due to the onset of an electronic contribution. Near room temperature, UO2 has a maximum in k which is apparently caused by the rapid decrease in specific heat below this temperature.

No abstract