Phase transition retrieval through in situ observations

Abstract

The aim of the investigation is to retrieve phase transitions upon heating moisture-containing materials using available observations of the specimen moisture state. Experimental data processing is treated as an approximation by the class of model functions that satisfy a given differential equation. The problem formulation differs from traditional parameter estimation and function reconstruction problems in that it is based on a relaxing regularization of a complex mathematical object. A given sample approximation is constructed under the condition that the measured data weakly capture changes in the state of an object. Such a case requires a high-resolution method for data processing that ensures the best consistency with observations. Regularization is applied to account for the ill-conditioned discretization of the mathematical model of the phenomenon and to ensure the approximation of a reconstructed function using a high-dimensional mathematical basis. The data processing of three moisture-containing materials demonstrates for the first time that thin-layer drying can involve a series of phase transitions. The latter causes a drop in the drying rate, which prevents moisture removal. Traditional postprocessing of the experimentally determined drying curve and its direct differentiation do not reflect mass transfer blocking and lead to a restricted interpretation of the experimental data. A proposed high-resolution processing of in situ observations, based on the application of the conventional thermal model, yields significant information on latent phenomena.