Cascade Measurements of Thermal Transpiration for Ar, N2, CO, and the Helium Isotopes

Abstract

A new experimental technique was employed to measure the thermal transpiration effect (thermomolecular pressure difference) for Ar, N2, and CO between 77 and 97°K and for helium-3 and helium-4 between 4.2 and 7°K. A cascade device consisting of seven large tubes connected top to bottom with capillaries was partially immersed in either liquid nitrogen or liquid helium while a suitable temperature difference was applied by heating the upper ends of the tubes and capillaries. Small pressure differences were established in each of the capillaries and the cumulative effect was large enough to be accurately measured with a differential capacitance manometer. The modified Weber–Schmidt theory was used to analyze the data and represented the argon data quite well. Some difficulty, which was traced to the coefficient of the temperature gradient, was encountered in obtaining a satisfactory theoretical explanation of the results for the other gases. We concluded that either an incorrect or oversimplified treatment of the temperature jump distance probably caused the discrepancy. Until further elucidation of this term, which also contains the translational Eucken factor, is forthcoming, use of this model for obtaining rotational relaxation times from thermal transpiration measurements is probably of limited value.