A new method for measuring the gravitational constant G is described. Preliminary measurements give G= (6.674 ±0.012) xio"" 11 N m 2 /kg 2 where the 0.012 represents 3 standard deviations. Furthermore there is reason to believe that with certain modifications of the apparatus and use of improved metrology techniques an increase in precision of at least one and probably two orders of magnitude will be obtained.

The original electromagnetic support developed in the late 1930's is a one-dimensional system. Servoed control is obtained in one direction and only inherent stability due to the field shape is obtained in the lateral directions. In this paper the more general problem of a three-dimensionally controlled support is treated theoretically. By virtue of making certain assumptions which seem reasonably close to practical feasibility two basic three-dimensional support schemes have been devised, in which ideally the three mutually perpendicular forces are uncoupled. The two arrangements are described and the theory is applied to predict support performance and to predict the amount of coupling to be expected due to deviations from the ideal system.

No abstract

The design of an electromagnetic balance using superconducting coils is reported. Both dc and ac coils are used to support aerodynamic models in a supersonic (Mach 3) wind tunnel and to simultaneously measure the forces acting on them along 3 orthogonal axes. Major design characteristics include: adoption of symmetrical coil arrangement to provide maximum space for the wind tunnel; 3 gradient-coil pairs capable of being driven between 0 and 350 A at a frequency of 30 Hz by specially designed power supplies; a vertical wind tunnel with a 6-in. test section located in the axial room-temperature access of a 250-liter liquid-helium Dewar. Results on ac losses for prototype gradient coils wound of three different superconducting materials are reported.

For some time this laboratory has been interested in the rotational magnetic losses in ferrites at low frequencies. Apparatus has been developed and successfully operated in which these losses are measured at frequencies of the order of 0.1 cps in fields up to about 5000 gauss and over a reasonably wide temperature range. It consists of a quartz fiber torsion pendulum (the fiber perpendicular to the field) operating in a vacuum and in a reasonably constant temperature enclosure. The principle involved is that of observing the decay of free oscillations of the system, and determining the torques due to the damped oscillatory system. By a process developed in our laboratory the samples are ground into spheres to accuracies of the order of 0.1% which can be improved it desired. In the absence of magnetic anisotropy of the sample, the method can, in principle, distinguish between Coulomb or constant torque losses (customarily assumed for rotational hysteresis loss torque) and loss torques proportional to the angular velocity (viscous). However, for all samples so far measured the anisotropy has been sufficiently large to prevent the disentangling of the Coulomb and viscous losses. (Qualitative indications of the anisotropy are obtained from the measurements and it is interesting that the losses and anisotropy seem to be proportional.) The procedure has been to assume the losses are of the constant torque type and determine them by making the best engineering fit with the decay curves.