Computer codes which model the pyrolysis of thermal insulators for in-depth temperature response are particularly sensitive to the kinetic parameters used in the code. The parameter values, which are evaluated from thermogravimetric analysis (TG), are, in turn, sensitive to the heating and gas flow rates in the TG sample environment as well as the composition of the gas. In this study, the effects of these experimental variables on computer code predictions and their correlations with experimental data are demonstrated. The insulators are urethane foams and were subjected to simulated and actual instantaneous fire test conditions. The Charring Material Ablation computer code was used to model insulator performance.Systems sensitive to heat that must be exposed to high temperatures require special consideration in the choice of their insulation, since margins of weight and size in the insulation are often critical to the overall system performance. It is desirable, therefore, to minimize insulation requirements. Computer codes have been developed to estimate the thermal requirements and to reduce the amount of confirmatory testing required to tassure acceptable performance. Unfortunately, in some cases poor agreement with large scale test data was obtained when the kinetic parameters derived from conventional laboratory data were used in the calculations [1][2][3]. To obtain satisfactory correlations between computations and observed test data, effective parameters derived from test data had to be substituted for parameters from laboratory data.This study addressed the importance of test conditions for data acquisition from thermogravimetric analysis (TG). These data were used to evaluate kinetic parameters which, in turn, were used to model the degradation of a urethane foam exposed to an instantaneous fire with a high heating rate.
