A Shuffled Complex Evolution (SCE) optimization routine was used to determine parameters based on some parameters being known (i.e., measured) and others being determined through optimization and different combinations of types of input data. The results indicate that material testing should be done at three different heating rates in a thermogravimetric analyzer (TGA) and at three different heat fluxes in the cone calorimeter. The TGA tests and two cone calorimeter tests are input into the SCE optimization routine to determine the remaining parameters. The third cone calorimeter test is used to validate the parameters determined using the methodology. The method provides parameters within 20 percent of the actual parameters input into the Fire Dynamics Simulator (FDS) virtual experiment models. This method is applied to Polymethyl methacrylate (PMMA) experimental data to prove effectiveness. (publisher abstract modified)
Methodology for Material Property Determination
NCJ Number
254475
Journal
Fire and Materials Volume: 43 Dated: 2019 Pages: 694-706
Date Published
2019
Length
13 pages
Annotation
Since predicting solid material pyrolysis requires numerous parameters/properties that are time consuming and difficult to measure, this study investigated combinations of testing protocols and parameters for optimization to determine a testing and optimization methodology that results in the accurate parameters.
Abstract