An improved mathematical model for the interpretation of the flow-through type diffusion test with influence of filterplates

A widely used method for the measurement of the diffusion parameters of a liquid saturated porous material is the so-called flow-through type diffusion test. To ensure its confinement and to avoid deterioration of the sample, the porous material is usually sandwiched between two filterplates. The influence of the filters is usually omitted in the interpretation of the experiments, assuming that the resulting error on the estimated parameters is small or at least tolerable. A mathematical model of the diffusion system including the filters is given together with its analytical solution. The errors resulting from the neglect of the filters is discussed in detail; errors larger than 40% are common and the value of the parameters is seen to depend on the duration of the test. An example is given where experimental data are interpreted with the complete model and with the filters omitted. Though in the latter case the parameter values are quite erroneous the mathematical fitting is equally good, illustrating that a good agreement between experiment and model does not necessarily establish the validity of the model. (author) 6 figs., 4 refs