OPTIMAL DESIGN OF RAMAN DIFFUSION EXPERIMENTS

Abstract

Model based diffusion experiments using Raman Spectroscopy is an ad- vanced technique to measure diffusion coefficients fot multicomponent and reactive mixtures with good precision and low measurement times. This work revolves around the designing of these Raman experiments. A new technique called model based experimental analysis can be employed which couples the high resolution measurements on one hand and analyzing the ex- periments using appropriate models on the other. Designing of the diffusion experiments aims at obtaining the precise diffusion coefficients by optimizing the design variables, which are the free settings of the experimental setup. But the irony of the situation is that, designing of these experiments itself requires good initial guesses of the unknown diffusion coefficients a priori. Poor guesses can adversely affect the design in the sense that the estimated diffusion coefficients are not collected with optimal precision. The main fo- cus of this project is to have a generalized design that is less sensitive to the choice of initial guesses of the diffusion coefficients. The work deals with binary, ternary and reactive mixtures.