Title: EXPERIMENTAL CHARACTERIZATION AND MODELING OF FLAME HEAT FEEDBACK AND OXIDATIVE PYROLYSIS FOR SIMULATION OF BENCH SCALE FIRE TESTS
Author: Conor McCoy
Date/Time: Thursday, August 5th | 9:30am
Examining Committee:
Professor Stanislav I. Stoliarov, Chair
Professor Mohamad Al-Sheikhly, Dean’s Representative
Dr. Richard E. Lyon
Professor Arnaud Trouvé
Professor Bao Yang
Abstract: Two important bench scale fire tests, the cone calorimeter test and UL-94V, were characterized experimentally to allow for predictions using a numerical pyrolysis solver, ThermaKin2Ds with pyrolysis parameter sets. Flame heat feedback was measured in cone calorimeter tests for several polymers to develop a generalized flame model. Flame heat flux was measured in the center and near one side and was found to be 11–23 kW m-2 and 32–49 kW m-2, respectively. Based on the difference in measured heat flux, a center zone and a side zone were defined and separate models developed. The final model was an area-weighted combination of the center and side zone simulations. Heat release rate data were predicted well by the final model. Ignition times for low irradiation were not predicted well initially but a correction was made to account for the effect of oxygen. The UL-94V test required characterization of the flame heat feedback but also of the burner flame (temperature, heat flux, and oxygen content). UL-94V tests were performed using polymers of different flammability ratings to evaluate the model; some samples had insulated sides to investigate edge effects. Additional UL-94V tests performed with an embedded heat flux gauge served to measure polymer flame heat feedback. All UL-94V tests were recorded on video using a 900-nm narrow-band filter to focus on emissions from soot for tracking flame length over time. Flame heat fluxes of insulated PMMA samples confirmed a previously developed wall flame submodel, while non-insulated PMMA samples had significantly greater heat fluxes; the wall flame submodel was scaled accordingly. Burner flame oxygen content was measured to be about 5 vol% and was found to enhance decomposition of two materials; oxidation submodels were then developed accordingly. Overall, the model predicted flame spread on insulated UL-94V samples reasonably well but significantly underpredicted the results on non-insulated samples. Discrepancies were attributed to burning and spread on the edges which were not modeled explicitly. Finally, given the importance of oxidation on predictions of ignition time, oxidative pyrolysis was studied both in mg-scale and gram-scale pyrolysis experiments. Kinetic parameters were first developed based on inverse analysis of TGA tests in atmospheres of varied oxygen content. Two models were developed: a surface reaction model and a volumetric model. Mass flux data from gram-scale gasification tests were used to evaluate the models. The anaerobic model gave the best predictions of mass flux for 15 kW m-2 gasification tests but the oxidative models gave better predictions for the 25 kW m-2 gasification tests. The volumetric model gives better predictions unless mass transport of oxygen is considered in which case, the surface model gives better predictions.