Digital Holography Measurement And Modeling Of Diffusivity For Binary System In Liquid Mixture

Diffusion in liquid is a mass transfer phenomenon caused by random molecularmovement. It is essentially happening in almost all unit operations handling liquidmaterials in chemical processes.. Diffusion coefficient, which reflects the speed ofdiffusion, is an important physical property that controls the phenomenon. How toacquire accurate diffusion coefficients is the core content in chemical engineeringresearch. We investigate both the measurement and the prediction theory of diffusioncoefficients in this work.Digital holography is a newly developing technique for diffusion measurement.It introduces CCD camera and computer techniques to traditional holographicmeasurement, which lead the capture, storage, transfer and post-processing of thehologram image to be more convenient. Digital holography has advantages such asreal-time observation, no flow field disturbance and high measurement accuracy. Inthis work, a digital holography system is devised and we describe in detail about thetheory of digital holography, the image processing and the method for extractingdiffusivities. Two methods for calculating diffusivity from the fringes directly areintroduced, namely the horizontal fringe method and the curved fringe method.0.33mol L-1aqueous solution of potassium chloride at25oC was used as areference system to check the experimental method. A comparison with literature dateshows that our method is reliable with good accuracy. Two methods are compared byemploying the measurement results of the ethylene glycol-water system. The curvedfringe method is showed to be more convenient and more accurate than the horizontalfringe method. Moreover, measurements were conducted at different concentrationsof acetone-water and other4binary liquid systems.A new model for predicting diffusivities of binary liquid mixture is developedbased on the local composition concept and Eyring’s Absolute reaction rate theory.The proposed model, extended from the Vignes equation, can predict the Maxwell-Stefan diffusion coefficients based on the diffusivities at infinite dilution，fromwhich the Fick diffusivities can be calculated. Fick diffusivities of nine binarysystems, five of which are measured experimentally in the present work and the otherfour are chosen from literature, are used to assess the proposed model. The proposedmodel was competed with the Vignes model as well as the other two models. The comparison of average relative deviation showed that the proposed model is superiorto the Vignes model for all systems and competed well with the other two models.