R. T. Johnson

PRINCIPLES AND METHODS OF DILUTION REFRIGERATION. II.

This paper is an extension of, and complementary to, an earlier paper by Wheatley, Vilches, and Abel. Some important fundamental questions which lead to degradation in the performance of a dilution refrigerator are considered: namely, convection on the dilute side and superfluid on the concentrated side. Experimental results are presented for a film suppressing still used in conjunction with a continuous-heat exchanger (several designs are considered) and a variable number (0–6) of step-heat exchangers of a copper-foil type with a high surface-area-to-viscous-impedance ratio. The first quantitative measurements of viscous heating on the dilute side are presented as well as measurements of thermal resistance on the dilute and concentrated side. Finally, the properties of the refrigerator under external heat load are considered.

Low temperature melting curve of He3

Measurements of the melting curve of He3 have been made from 3 m° K to 40 m° K, after cooling the liquid-solid He3 mixture by the adiabatic solidification method. The slope of the melting curve shows that at the lowest temperature there is a substantial departure of the solid He3 entropy fromR ln 2. This effect is due to nuclear spin ordering. Temperatures below 3 m° K were obtained during compression. By theoretically estimating the melting curve below 3 m° K it is possible to estimate temperatures from pressure measurements. By such means it is estimated that Kelvin temperatures of 1.5 to 2 m° K were actually achieved.

Heat flow in superfluid 3He

The thermal resistance in both superfluid phases of 3He has been measured at 20.0 and 29.6 bar in zero magnetic field. Heat conduction in 3He-B is shown to be primarily hydrodynamic, and a regime of reproducible heat flow behavior in the A phase is reported. The viscosity of each phase as a function of temperature is calculated using an equation of the two-fluid model, and critical velocity effects are discussed.

Effect of a magnetic field on the melting curve of3He

The melting curve of3He was measured below 25 mK and in magnetic fields of 2.12, 6.36, 31.8, and 63.6 kG. The compressional cooling technique was used with a capacitive pressure gauge and a54Mn-in-iron γ-ray anisotropy thermometer. A depression was observed in the melting curve, indicating nuclear spin ordering in the solid3He. The effect is larger in magnitude and develops more rapidly with field than the theoretically expected results.

Bulk nuclear polarization of solid3He

Measurements are given on the bulk nuclear spin polarization in a liquid-solid3He mixture cooled by compressional cooling to below 5 mK in a magnetic field of 54.5 kG. Owing to the low Pauli spin susceptibility of liquid3He, the observed polarization is primarily due to solid3He. A maximum average nuclear polarization of 47% was observed, although the corresponding solid3He polarization is believed to be higher. Our novel detection system, using a dual directional coupler for cw NMR, is a simple and versatile means of working in the awkward frequency range around 180 MHz. We also report transient heating measurements in the3He system which indicate that the internal thermal equilibrium time in bulk solid3He on the3He melting curve appears to be quite short (less than 5 min) at these temperatures. One type of transient measurement is complicated by the dramatic effect of the contribution of the3He nuclear magnetization to the total local magnetic field. This contribution is considered via a simple model.

Effect of a magnetic field on the melting curve of3He—Addendum

An additional analysis of our low-temperature and high magnetic field measurements of the3He melting curve is presented, based on the assumption that the highest field (63.6 kG) theoretical melting curve is correct rather than the lowest field curve as assumed in the original paper. Between31.8 and63.6 kG there is reasonable agreement between experiment and theory, although a discrepancy remains at low fields that we assume is to be attributed to inaccuracies in the reduction of γ-ray anisotropies to temperatures. It is also shown that the measurements are not sufficiently precise to observe the effect seen in the(∂P/∂T)V,H measurements on solid3He by Kirk and Adams.

Qualitative changes in the self-diffusion coefficient and T2 for solid 3He at very low temperatures

Abstract When solid 3 He is cooled by compression, over a small temperature range at the lowest temperatures both D and T 2 drop substantially from those nearly temperature-independent values they have at higher temperatures.