Recently, it becomes very important to measure low pressure with fine accuracy. But the conventional vacuum gauges have much errors : usually several tens and sometimes several hundreds per cent. And many confusions have occured from this inaccuracy.
To eliminate these confusions, we have started to make the standard gauge and to establish standard calibration method three years ago. The standard gauge capable of the calibration of ionization gauges in the range of 10^-510^-2mm of mercury is a large McLeod gauge whose volume is 1, 000cm³ and whose capillary diameter about 1mm. Calculated error of this gauge is about one per cent at pressure of 10^-5mm of mercury, but the experimentally obtained error is increased to 2 per cent at 10^-4mm of mercury. This reason is considered mainly due to the sticking property of mercury thread to clean glass surface in the high vacuum.

抄録全体を表示

Preliminary work on a spectral analysis type leak detector is described.
The detector consists of a discharge tube, a spectrometer, a photomultiplier and an amplifier. An electrodeless discharge tube which is attached to the vacuum system is operated by a high frequency current of about 100 megacycles modulated by 60 cycles. The emitted light from the discharge tube is dispersed by a constant deviation prism spectrometer and received by a photomultiplier 1P 21. The audio frequency parts of the signals from 1P 21 are amplified and rectified so as to operate a D.C. milliammeter.
When each small quantities of air, hydrogen and helium is introduced to the vacuum system through an adjustable leak valve, several peaks corresponding to the spectral lines characteristic of the gases appears in the meter. In the present experiment, the practically detectable leak quantities are about 10⁴mmHg 1/sec for hydrogen and 10⁵mmHg 1/sec for helium.
Comparing with the mass spectrometer type leak detector, the new detector has the following features :
1) No need of vacuum system in the detector,
2) Portable and durable constructions,
3) Simple circuit and prompt operation.

抄録全体を表示

In the measurement of the pump speed of an oil diffusion pump by Dayton's method, the speed measured with a test dome of a larger diameter than that of the pump mouth shows a value smaller than the one obtained with test dome of the same diameter as the pump mouth.
The reason for it is considered to lie in that the former conductance is smaller than the latter, when a distance between the pump mouth and the measuring position is smaller than a mouth diameter.
We measured pressure distribution in an axial and a radial direction in a test dome and determined the conditions in the test dome for speed measurement.

抄録全体を表示

The gastightness of the contact interstice of polished surfaces of metal is studied by means of the leakage of an indicator gas through the interstice. The clearance of interstice δ, ranging ordinarily from 1 to 10μ, is narrowed by a load W, being increased from 0 to 1500 Kg/cm² step by step. The half of δ, ranging from 0.5 to 5μ, is the thickness of the outermost layer, which stands against our experiment, of the polished surface of metal. The estimation of δ on the diagram of log δ versus W is accurate within several percents of the magnitude of δ in the case.
The polished layer of metal is in an expanded state, and is highly plastic. When such a layer is compressed, several similar modes of deformation start side by side in the layer, and end individually in the loading, while increasing compression. Accordingly the resulting figure of log δ versus W usually consists of several segments of straight line. The excess of the energy is stored in the polished layer, being split into a spectrum, corresponding to the constituent deformations mentioned above. For eamople :
ΔF_a=2.4×10^-5, ΔF_b=2.2×10^-4, ΔF_c=4.9×10^-4, ΔF_d=3.83×10^-3Cal/cm²., and the sum ΣΔF_i=4.56×10^-3Cal/cm². The last is estimated to be equivalent to 31.5 Cal/cc of the polished layer or to 4.1 Cal per gram of iron.

抄録全体を表示