BUBBLE GROWTH IN VISCOUS LIQUIDS

Abstract

The removal of volatile contaminants from molten polymers is known as Polymer Devolatilization (DV) process, during which superheat or vacuum are often applied as common practices to form foam to effect separation. The growth of bubbles containing vapor of the residues is a key step of foaming DV, which has become a focused research subject in aim to help the understanding and improvement of DV operation. In this work, a theoretical model is established first to analyze the growth kinetics of an ideal gas bubble in an infinite quiescent Newtonian liquid. Our simulation takes into account not only mass transfer, but also momentum forces. Computer programs are designed in order to solve all the control equations simultaneously in a numerical form. Approximate analytical solutions are also developed to characterize the bubble's basic features under different physical conditions. The most significant result we discovered is the rapid pressure build-up inside the bubble under strong driving forces right after the bubble's birth. Since very high pressure can invalidate the ideal-gas assumption, the non-ideal-gas effect of the bubble component is further studied. Because van der Waals state equation is capable of predicting phase change, it is employed to replace the ideal-gas state equation that is often used in earlier traditional models. We make a second innovation by introducing Flory-Huggins theory to relate the bubble pressure with the solvent fraction in the liquid. For it is valid for the whole range of the solvent fraction, which is more realistic than the widely used Henry's law. With the coupling of van der Waals equation and Flory-Huggins theory, possible wrong physical situations where the pressure inside the bubble exceeds the solvent's saturation pressure are avoided. A sample polymer devolatilization system is proposed to show the superiority of the non-ideal-gas simulation mechanism over the ideal-gas one, before the final summary of the advantages and shortcomings of our bubble growth model.