Modeling and simulation of linear thermoplastic thermal degradation

Access full-text files

Date

2012-05

Authors

Bruns, Morgan Chase

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Thermal degradation of linear thermoplastics is modeled at several scales. High-density polyethylene (HDPE) is chosen as an example material. The relevant experimental data is surveyed. At the molecular scale, pyrolysis chemistry is studied with reactive molecular dynamics. Optimization is used to calibrate several pyrolysis mechanisms with thermogravimetric analysis (TGA) data. It is shown that molecular scale physics may be coupled to continuum scale transport equations through a population balance equation (PBE). A PBE solution method is presented and tested. This method has the advantage of preserving detailed information for the small species in the molecular weight distribution with minimal computational expense. The mass transport of these small species is modeled at the continuum scale with a bubble loss mechanism. This mechanism includes bubble nucleation, growth, and migration to the surface of the condensed phase. The bubble loss mechanism is combined with a random scission model of pyrolysis to predict TGA data for HDPE. The modeling techniques developed at these three scales are used to model two applications of engineering interest with a combined pyrolysis and devolatilization PBE. The model assumes a chemically consistent form of the random scission pyrolysis mechanism and an average, parameterized form of the bubble loss mechanism. This model is used to predict the piloted ignition of HDPE. Predictions of the ignition times are reasonable but the model over predicts the ignition temperature. This discrepancy between model and data is attributed to surface oxidation reactions. The second application is the prediction of differential scanning calorimetry (DSC) data for HDPE. The model provides detailed information on the energy absorption of the thermally degrading sample, but the literature data is too variable to validate the model.

Description

text

LCSH Subject Headings

Citation