Use of a response surface methodology to model thermal decomposition behavior of polyurethane.

In this study, thermogravimetric (TG) curves of a synthetized renewable polyurethane–urea were modelled in intermediate heating rates from those experimentally tested. The response surface methodology (RSM) was used to optimize the tests by creating new TG curves. Flynn–Wall–Ozawa kinetic model was applied in the theoretical curves and re-applied in the new theoretical ones. The predicted curves showed similar curve format, but different kinetic results compared to the experimental ones (Ea = 170.56 kJ mol−1 and A = 4.11E10 s−1 for the experimental curves, and Ea = 84.71 kJ mol−1 and A = 0.02 s−1 for the predicted ones). Also, Pearson correlation coefficient was plotted in function of the heating rate, mass loss at 5%, the three peak temperatures and mass residue, showing that the mass loss has no significant influence on the other parameters tested. Considering the heating rates and the peak temperatures, the p values were 0.11, 0.078, and 0.056 for the Tp1, Tp2, and Tp3, respectively. This means that there are 89%, 92.2%, and 94.4% of chance that the results are not due to random chance. As main conclusion, the poorer correlation among the variables of the predicted curves using the Pearson correlation reflected in distinct Flynn–Wall–Ozawa parameters. [ABSTRACT FROM AUTHOR]