Habilitation Thesis
DOI
https://doi.org/10.11606/T.3.1994.tde-11082022-081900
Document
Author
Full name
Reinaldo Giudici
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 1994
Committee
Brunello, Giovanni (President)
Biscaia Junior, Evaristo Chaubaud
Mendes, Mario de Jesus
Mori, Milton
Nascimento, Claudio Augusto Oller do
Title in Portuguese
Modelagem matemática do processo de modificação química de polímeros.
Keywords in Portuguese
Polímeros (Química orgânica)
Abstract in Portuguese
Title in English
Mathematical modeling of the chemical modifications of polymers.
Keywords in English
Polymers; Mathematical modeling
Abstract in English
Mathematical modeling of the chemical modification of polymers by free radical mechanisms was the main scope of this work. Simultaneous reactions that place in this process include scission, crosslinking and grafting. An integro-differential equation describes the variation of the molecular weight distribution (MWD) with conversion. Numerical solution was done by discretizing the MWD in a number of nodal points, and solving the system of differential ordinary equations by predictor-corrector Adams method. The distribution is interpolated using a cubic spline polynomial, that is used to evaluate the integrals. Some discrepancies are found to occur with the numerical solution near the gel point. It was studied the degradation of polypropylene, a process in which random scission is the main reaction. By fitting the model to literature data, process parameters were estimated. The conventional approach to model scission considers that there are two ways to CUT a chain to make a given product. This is exactly correct only for linear chains. The rigorous prediction of scission of branched chains is an unsolved problem in the literature. In order to study this process, two different approaches were applied and discussed. Stars and combs were used as idealized structures in the simulation with both approaches. The first approach is a simplified, deterministic one where it is assumed a simple, idealized chain morphology for all chains in the population. The second approach uses Monte Carlo methods and it is much more realistic as long as it is able to keep track of the structure of each chain during the process.