Standardising Design Fires For Residential and Apartment Buildings: Upholstered Furniture Fires
Thesis DisciplineFire Engineering
Degree GrantorUniversity of Canterbury
Degree NameMaster of Engineering in Fire Engineering
This purpose of this research was to develop a credible set of furniture design fires for residential/apartment buildings and determine a methodology for incorporating compartment effects in design fires. Design fires can be defined using various outputs, the most important being the HRR profile, and depending on the application the following may also be relevant: · Smoke production rates · Soot yield · Species production rates · Temperature profiles · Visibility · Heat fluxes · Mass loss rate of the fuel · Flame spread There were three phases to this project: The first phase of this project was a comprehensive data and literature review to determine the amount of experimental data available and commonly accepted burning characteristics for upholstered furniture; armchairs, 2-seater sofas, 3-seater sofas, beds and bedding assemblies, and commonly accepted burning characteristics and compartment effects. A large proportion of the review provided only qualitative guidance for design fires. In the second phase the data collected during the review was collated and used to quantitatively analyse key fire characteristics. These were · peak HRR, · time to peak HRR, · growth rate, · total heat released and · maximum CO/CO2 ratio. A methodology was developed to statistically analyse experimental data using BestFit, and where there was sufficient data the 98th percentile of the statistical analysis was used as a quantitative guide for furniture design fires. Similarly, compartment effects were incorporated into the design fires by analysing and comparing the experimental data from free burn and room burn tests of the same furniture item. The same statistical analysis was used to determine likely changes in the key fire characteristics mentioned above. A methodology for determining design fires for upholstered furniture was devised, however the small number of data sets available for analysis meant the quantitative results were only indicative. The third phase was to attempt to model furniture fires using FDS, which determined that at the time of this project, FDS was not capable of modelling simple furniture fires accurately. The simulation results varied significantly from the experimental results and a number of limitations were identified. Therefore FDS should not be used to create design fires using the heat of combustion method, which relies on the users’ definition of material properties.