Citation

McCoy CG, Stoliarov SI. Fire Mater. 2022; 46(6): 905-918.

Copyright

(Copyright © 2022, John Wiley and Sons)

DOI

10.1002/fam.3038

PMID

unavailable

Abstract

Previously, a model of the UL-94V test combining a comprehensive two-dimensional pyrolysis model with an empirical heat feedback model of the burner and polymer flames was developed and tested using extruded poly(methyl methacrylate) (PMMA). PMMA samples with and without insulated sides were used to investigate edge effects: the propensity for the flame to ignite and spread at the edges. Flame growth dynamics were quantified by processing video recorded at 900 nm wavelength. UL-94V tests were simulated using a numerical solver, ThermaKin2Ds; flame length was predicted well for the insulated samples but not for the non-insulated samples, likely due to edge effects. In this study, the model was evaluated against UL-94V tests of seven other polymers of different UL-94V ratings, using both insulated and non-insulated samples. Thermogravimetric analysis was performed in 5 vol% oxygen to check the materials' sensitivity to oxygen; only high-impact polystyrene was affected, for which a surface oxidation expression was developed. For the materials that sustain flame, predictions of flame spread for insulated samples were good after dripping and soot accumulation were considered, although initial flame length was underpredicted for two materials. Predictions for the non-insulated samples were worse for all materials, likely due to edge effects. For the materials that did not sustain flame, the model predicted extinction correctly in 50% of them. It was unable to predict extinction due to gas-phase flame inhibition since such effects are not included in the model. However, for these materials, predicted flame lengths before extinction were excellent.

Language: en

Keywords

concurrent flame spread; edge effects; flame heat flux; oxidative pyrolysis; pyrolysis modeling; ThermaKin; UL-94