Reconstruction of tall buildings fires for thermal response analysis

Abstract

In recent years, fires in tall buildings have become more frequent, which costs billions of dollars each year and the loss of many human lives. Uncontrolled fires caused three supertall steel framed structures to collapse completely in New York following the terrorist attack on 11 September 2001 resulting in thousands killed, including hundreds of firefighters; the horrific fire that spread rapidly across the entire facade of the Grenfell tower cost 72 lives; and earlier the same year 20 firefighters were killed in the collapse of the Plasco Tower in Tehran. Despite such extreme events that continue to occur periodically, no well-defined methodology exists to reconstruct both the fire and structural behaviour and carry out a comprehensive forensic investigation of building fires. Using currently available tools and data, this thesis proposes a novel and robust methodology to reconstruct the fire for the forensic assessment of tall buildings that can guide the structural and fire safety engineers to improve the building fire-safety and life-safety designs. The collapse of the Plasco Building is assessed by employing the proposed framework. The fire scenario that led to the collapse of the Plasco Building is reconstructed. A well-defined sequence of events is presented to establish a rational fire timeline in conjunction with structural damage that occurred in the building during the event. In the past century, exhaustive studies have been conducted to understand the structural response to different fire scenarios. Various fire models have been proposed so far, such as standard time-temperature curve, highest temperatures in a compartment, and localised fires to define the fire scenario in a large open compartment. Chapter 2 reviews and presents the limitations and applicability of some widely adopted fire models to understand their suitability for structural fire assessment. This chapter also presents the fire-models developed in the advancement of the performance-based engineering (PBE). The recent trend toward PBE approaches is beginning to require even more accurate representations of fires and consequently more realistic boundary conditions for fire exposed structural components, leading to greater use of computational models. More sophisticated computational fluid dynamics (CFD) based fire models and analytical models to present travelling fires have emerged in the last two decades. Travelling fire models produce the most realistic building fire models at low cost but they are only meant to provide a simple method for estimating design fire scenarios for performance based engineering (PBE). CFD provides much better resolution (spatial and temporal) of the thermal boundary conditions over structural surfaces. However, it is a challenging task to couple a CFD model with finite element methods (FEM) for simultaneous or coupled simulations of both the fire and the structural response. This thesis focuses on how the most important prerequisite information for a reliable structural analysis is achieved i.e., getting the realistic thermal boundary conditions for structural analysis and how to automatically provide to an FEM package for heat transfer and structural response analysis. A generic middleware is developed in Chapter 3, which interfaces FDS and OpenSees to enable fully integrated simulation of fire scenario, heat transfer, and structural response. Using the middleware, to determine the structural response to fires, an open-source package (OpenFIRE) is developed. OpenFIRE is an only open-source and licence-free computational framework that allows the structural fire community to customise and modify according to client requirements and rigour of the analysis desired in the context of PBE. OpenFIRE could enable even small consultants to undertake large projects requiring PBE solutions against the fire hazard for existing and new buildings. OpenFIRE is a robust tool for the analysis of structural response to realistic fires (such as localised, spreading or travelling fire phenomena), which should help enable performance based engineering and design of structures to real fires. Forensic investigators can use OpenFIRE to carry out structural analysis of a building that has been damaged in a fire incident. In this thesis, OpenFIRE is used to obtain thermal boundary conditions for the fire in the Plasco Building, Tehran (2017) by modelling the real fire in that building using CFD. A methodology to carry out forensic investigations of tall building fires is presented in Chapter 4. It facilitates generating a conclusive fire timeline by collecting the information from multiple resources such as visual evidence and interviews. Chapter 4 investigates the collapse of the Plasco building using the proposed methodology. Using evidence, the fire scenario that led to the collapse of the Plasco Building is estimated, , which is used to calibrate the CFD models. For performing a forensic investigation of a fire accident, it is critical to understand the fire dynamics involved in a compartment fire. Various codes and standards explicitly mention the fire load for an occupancy type; however, the effect of fuel distribution is scarce in the literature. In Chapter 5, a qualitative study is performed to understand the effect of fuel distribution on the fire spread behaviour in a building. It helped reconstruct the probable fire scenario that led to the rapid spread of fire in the Plasco Building, which eventually collapsed within three and half hours after the fire broke out. Chapter 6 uses CFD to reconstruct the empirically estimated fire spread of the Plasco building presented in Chapter 4. The vertical and horizontal fire spread in the Plasco Building is reconstructed using CFD fire modelling and calibrated with the evidence library generated in Chapter 4. The thermal data obtained from the calibrated fire simulation is transferred to the OpenSees using OpenFIRE (developed in Chapter 3). The thermal data can be used as input boundary conditions to simulate the structural response and collapse of the building.