Date of Award

3-1972

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Materials Science and Engineering

Major Professor

J. E. Spruiell

Committee Members

Earl McHayes, E. E. Stooksbury, Howard L Davis, John W. Prader, S. L. Oliver

Abstract

Thermal conductivity, electrical resistivity, and thermoelectric. power measurements were made on thorium nitride (ThN), (Th-2% U)N, and (Th-5% U)N; electrical resistivity and thermoelectric power measurements were made on uranium nitride (UN). Results from the experiments indicated that ThN behaves very nearly as a pure metal in its transport properties. Thermal conductivities, λ, were high, electrical resistivities, ρ, were low, and thermoelectric power, S, measurements were nearly zero. Uranium nitride, which antiferromagnetic below 50°K and has a lower thermal conductivity, had a much higher electrical resistivity and a large positive thermoelectric power.

The thermal conductivity of ThN is much higher than nuclear fuels currently being considered for use in liquid metal fast breeder reactors. At nuclear reactor operating temperatures of interest, λ for ThN is approximately 20 times that of UO2. Even when alloyed with UN, ThN retained a thermal conductivity greater than any other ceramic fuel currently being considered. Calculations were presented which illustrate the potential of ThN alloyed with UN or plutonium nitride (PuN) as fuels in proposed reactor concepts. The results suggested ThN would have a significant economic advantage and a predicted fuel performance which exceeds other fuels being considered.

The electrical resistivity measurements showed ThN and the (Th-U)N alloys to have a linear dependence with temperature above about 100 °K. Below 100 °K, the slope changed to near zero at 0 °K. Small additions of UN to ThN of 2 and 5% caused a very large increase in ρ. This probably resulted because the antiferromagnetic UN exhibits a large ρ. In addition to changing the residual resistivity at 4.2 °K, dρ/dT was reduced by alloying.

The thermoelectric power of ThN is very small. However, alloying with UN, which has a large value for s, caused a significant change in the S of (Th-U)N alloys.

The data treatment permitted a separation of λ into a lattice conductivity, λL, and an electronic contribution to the thermal conductivity, λe. The results showed both of these contributions to be significant over the temperature range studied. The ability to separate the components accurately allowed a prediction of λ at temperatures greatly exceeding the range of actual experimental measurements. The data treatment also resulted in a calculation for the Lorenz function which was found to be very near the Sommerfield value at temperatures of 300 °K and above.

A significant aspect of the study was the development of a technique to make high quality samples. None of the standard fabrication techniques were satisfactory, so a zone melting technique was developed which converted the metal directly into the nitride. This technique produced high-purity, high-density samples approximately 100 times as fast as any of the other more conventional techniques. The fabricated samples were nearly 100% dense with very Large grains of approximately 0.3 cm diameter. High-density, large-grained samples would be expected to show superior in-reactor performance to either small-grained or low-density samples.

Another important aspect of the study was the use of the PDP-8 computer to take and analyze some of the data. By using the computer system, the data were taken approximately 1000 times faster in the case of high temperature ρ measurements. This is very important for a material such as UN which has a large value for Sand exhibits significant Peltier heating on reversing the electrical current. In the case of UN, this was the only way accurate data could be obtained above 1200 °K.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS