Date of Award

2007

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Chemical Engineering

First Advisor

Simant R. Upreti

Abstract

Diffusion in molten polymers far above the glass-transition temperature is characteristic of many industrial processes such as polymerization, monomer reclycling, stripping drying coating and foaming. Many of these systems of pracical importance exhibit a strong dependence of dissusion coefficients on concentration and temperature. These operations involve concentrated solutions of polymers and solvents, which are far removed from the dilute region where theoretical advancements have been most significant. The design and optimization of these applications requires concentration dependent diffusivity data, which are scarce at present.

In this work, the calculus of variation is used to establish the necessary conditions of the concentration-dependent diffusivity for a unidrectional distributed parameter model such that the model-predicted mass of absorbed gas in polymer matches with its experimental counterpart. A computational algorith is implemented to solve the model, and obtain the diffusivites of carbon dioxide gas in low-density polyethylene (LDPE) melt, in the rage of 352 to 1232kPa, at 120°C and 130°C. The optimal diffusivities versus concentraion in the polymer medium. The peak diffusivity of carbon dioxide in low density polyethylene melts for the above temperature and pressure range varies between 3.04x10-9m²/s to 4.56x10-9m²/s. The above results obtained are evaluated from their sensitivity with respect to maximum expected experimental variation in saturation weight fraction of the gas. The sensitivity of diffusivity to change in above system paramters is maximum as its peak value and is less than 2% with respect to its base value.


Download

Share

COinS