A top-down approach to understand the fire performance of building facades using standard test data
File(s)
Author(s)
Bonner, Matthew
Type
Thesis
Abstract
Designing the outer walls of a building – its facade – is a complex, multi-objective problem. While some design objectives, like weight, are easily quantified others, like fire safety, are harder to assess, making it difficult to compare facades in that objective. A move to more complex facades with novel materials has led the number of facade fires to quadruple over the last 3 decades. Current fire safety literature has no way to accurately predict the performance of a facade from its individual components, therefore the only way to test whether a facade system is safe is to run a large-scale fire test. Thousands of these tests are run internationally each year, subjecting different facades to different fire scenarios depending on the country, but data from these tests are locked away, missing an opportunity to learn from this untapped source of knowledge. This thesis taps into that source for the first time by presenting a unique database of 384 facade tests, named KRESNIK. We analysed this database using a top-down, statistical approach, and found that facades with combustible cladding and a cavity generally performed worse in these tests. These trends were investigated further in a parametric series of 20 intermediate-scale experiments on facades with different cladding and insulation materials, with and without a cavity. These experiments were analysed using a novel method to measure the heat released through visible flames in a fire from regular camera footage. We refer to this quantity as visual fire power. KRESNIK was also used to compare, for the first time, the behaviour of similar facades tested using different test standards, finding that different standards could not agree on how to rank the same facades. This work demonstrates the power of a top-down approach to tackle fire safety using data that already exists.
Version
Open Access
Date Issued
2020-12
Date Awarded
2021-07
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Rein, Guillermo
Vaidyanathan, Ravi
Sponsor
Engineering and Physical Sciences Research Council
Arup
Grant Number
EP/N509206/1
Publisher Department
Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)