AVS 46th International Symposium
    Vacuum Technology Division Tuesday Sessions
       Session VT-TuM

Paper VT-TuM9
A Review of Thermal Transpiration Corrections in Capacitance Diaphragm Gauges

Tuesday, October 26, 1999, 11:00 am, Room 610

Session: Total and Partial Pressure Gauging
Presenter: J. Setina, Institute of Metals and Technology, Slovenia
Correspondent: Click to Email

Capacitance diaphragm gauges (CDGs) play important role in vacuum metrology because of their accuracy and resolution. They are widely used in calibration laboratories as reference and transfer standards for calibration of other pressure measuring devices in the pressure range from 0.01 Pa to 100 kPa. CDGs are inherently susceptible to ambient temperature variations. Operating CDGs at stabilised elevated temperature greatly improves the zero stability and enables relative resolution of few ppm of gauge full-scale but also introduces nonlinearities, which are known as thermal transpiration (TT) effect. In applications, where the uncertainty of measurements must be reduced to less than one percent, the response of CDG has to be corrected for TT effect if measured pressure is below 100 Pa. TT effect had been studied in detail by Takaishi and Sensui,@footnote 1@ and their empirical formulas are widely accepted for corrections of CDG measurements. In the present paper we will address the accuracy of this formulas which were published four decades ago. We have done some accurate calibrations of CDGs in the pressure range from 0.01 to 133 Pa in four gases: N@sub 2@, Ar, He and H@sub 2@. For He we found discrepancy of our data from Takaishi and Sensui empirical formula as large as 0.3 % in the range of pressures from 0.5 to 5 Pa. Scaling of thermal transpiration curves for different gases with the inverse of Knudsen number was also observed. This has not been reported by other authors. Vacuum metrology greatly advanced since the work of Takaishi and Sensui. The improvement in measurement accuracy at primary laboratories in the past ten years allows much more precise measurements in the range from 0.01 to 100 Pa. Therefore we suggest a thorough re-evaluation of the TT effect that should include a variety of gases with which CDGs are commonly used as well as the influence of momentum exchange of gas molecules on the surface of the tube that connects a CDG to the vacuum system. @FootnoteText@ @footnote 1@ Takaishi T., Sensui Y.,Trans.Faraday Soc., 1963, 59, p. 2503