Prediction of time lag in vacuum tubes and its implications for characterization of membranes in constant volume systems: Effect of resistance-free accumulation tank

Abstract

In this project a "time limitation" of the previous analysis has been overcome by allowing the diffusion coefficient downstream from the tested medium to vary with pressure, describing this relationship using the empirical model of Knudsen. This required a numerical solution of the governing partial differential equation, which has been done by applying a finite difference scheme. The theoretically obtained pressure responses have been compared with the experimental pressure responses in two different flow cases, a constant flow and a time-dependent flow at a tube entrance. The former case was realized by using a low-flow mass flow controller as a flow source. The latter case was realized by using a homogeneous polymeric film, whose one side was instantaneously pressurized in a step-wise manner (time lag technique). The theoretical model slightly underestimates the resistance effects. This has been attributed to an unavoidable presence of a dead volume in the experimental system, i.e. the volume associated with valves, pressure sensors, bypass tubes, etc. A procedure for correcting the experimentally determined transport coefficients of gas in a membrane for the resistance effects has been developed and demonstrated. (Abstract shortened by UMI.)