Comparison of methods for thermal analysis: Application to PEEK and a composite PEEK+CF
Florian Hache,Michael Delichatsios,Talal Fateh,Jianping Zhang
DOI: https://doi.org/10.1177/0734904115584154
2015-05-01
Journal of Fire Sciences
Abstract:This article applies and critically compares essentially all methods (based on Arrhenius temperature dependence and power law reduced mass) to deduce the degradation parameters of polyether ether ketone (PEEK) and its carbon fibre composites, PEEK+CF, from thermogravimetric analysis experiments in nitrogen. The methods used include classical (and variations) analytical approximate methods, a method applying the genetic algorithm for optimization and the Distributed Activation Energy Model, usually not discussed with the other methods. The analytical methods are fast and easy to implement but they are limited to degradation kinetics having a single peak or individual separated peaks in the mass loss rate. By contrast, the genetic algorithm gives good results if a postulated mechanism of several Arrhenius reactions of arbitrary order is considered. Moreover, the Distributed Activation Energy Model provides the possibility to assess the assumption of overlapping first-order reactions having various activation energies. The degradation of PEEK and its composite is basically characterized by two regimes, one fast degradation regime (570 °C–630 °C approximately) and one much slower degradation regime (630 °C–800 °C approximately) at a mass degradation rate with time being simply proportional to the heating rate, a very remarkable new result by itself. However, the focus of the modelling in this work lies in the first fast degradation regime. It is shown that the Distributed Activation Energy Model and the genetic algorithm can simulate well the experimental results noting, however, that the Distributed Activation Energy Model is based on more physically determined activation energies. By contrast, the classical methods (Kissinger and modified Ozawa methods) are not able to reproduce the mass loss rates in the initial fast degradation regime (570 °C–630 °C) of pure PEEK and PEEK+CF composite. Although a physical explanation of the degradation mechanisms is desirable, the end game is to find out whether a method even not including the whole truth about a process is yet capable of reproducing the measurements accurately.
materials science, multidisciplinary,engineering