Computer Modeling - Term Paper Example

FIRE MODELLING Potential applications and limitations of computer modeling to fire and explosion investigations Table of Contents Contents 1 Methods and Types of Fire Modelling 3 1.1 Fire Modelling 3 2 Types of Computer Fire Modelling 5 2.1 Zone Models 5 2.2 Field Models 7 3 Application of Computer Modelling in Field Fire Investigation with Proof 9 3.1 Computer-based Fire Models for Fire Investigation 9 3.2 CFD Model Discovered the Cause of the 1987 Kings Cross Underground Fire 11 3.3 FDS and LS- DYNA for Simulating World Trade Centre Fire 13 3.4 FDS physical simulation of Station Nightclub Fire 15 4 References 16 1 Methods and Types of Fire Modelling 1.1 Fire Modelling Types of Fire Modelling include physical and mathematical representations of fire-driven processes . The physical model according to often involves full-scale reduced scale reproduction of actual compartment or building where materials are burned to simulate fire development. Moreover, physical models enable researchers to acquire simulated compartment temperature, smoke movement and accumulation, ventilation effects and others. Figure 1 - Types of Fire Models Mathematical fire models on the other hand are divided into probabilistic or stochastic and deterministic models. Probabilistic models are highly dependent on fire statistics and often come in event tree formats where transition from one state to another state of fire development is demonstrated. The equation L/D = 0.2 (Q2/5 /D) is a simple mathematical model can describe the relationship shown in Figure 2 where L is average flame height, D is diameter of the fire, and Q is heat release rate measured in kW . Figure 2 – Simple Mathematical Model Deterministic models (i.e. computational models) are more advanced as they can provide detailed predictive information on specified fire scenarios using physics and chemistry based mathematical equations . Deterministic mathematical fire models can range from a simple one-line correlations of data to complex models requiring dozens of computers and weeks of computing to complete . The rationale behind this model according to is that the course of a fire is fixed by different environmental variables. For this reason, the physical conditions fire scenario involving fuels and their arrangement inside the compartment, compartment shape and dimension, existence or absence of fire protection, and others determines fire progress and outcome. Mathematical fire models are again classified into field and zone models . The field models divide the compartment into three-dimensional grid and compute compartment conditions at each point on the grid. In contrast, zone models divide compartment into two layers of hot and cooler temperature . The first zone model was RFIRES published in 1981 followed by Harvard series of models, FIREFORM (origin of FPETOOL), and others. 2 Types of Computer Fire Modelling 2.1 Zone Models Computer fire modelling was first developed in the 1970s and was first applied to solve fire engineering problems in the late 1980s. Today, computer fire modelling is being applied to forensic investigations risk assessment, life safety, smoke movement and detection, evaluation of sprinkler performance, structural behaviour, and more . Zone models are a commonly used fire model as they can predict fire behaviour based on gases movement in the upper and lower zones. Examples of zone models are ASET-BX or Available Safe Egress Time-Basic, Harvard Code, FIRST, FPETool, Hazard – I, CFAST or Consolidated Fire and Smoke Transport. Zone models are a physical and deterministic fire model that predicts fire development inside the enclosure through conservation of mass, momentum, and energy calculations . Figure 3 –Schematic representation of assumed two layers in zone modelling As a computer-based fire model, zone models can predict vertical and horizontal movement of mixed smoke layer from the fire plume. For this reason, zone models according to , is often employed to perform smoke management in building structures. As a computer-based model, the advantages of zone models are the speed, ability to represent a large structure, and low memory requirements . However, since zone models are relatively simple and model fires occurring in a rectangular rooms with flat horizontal fuel beds, vertical walls, ceiling, and furniture, irregularly shaped compartments and movements of air and gases coming from secondary sources may affect the accuracy of zone models prediction. Moreover, although zone models simulation of fire plumes, airflows from vents and heat transfer provide reasonable results, they cannot provide detailed spatial; information on the fire environment . CFAST is a computer-based two-zone model developed by NIST for calculating gas temperatures, smoke layer height, pressure, and so on . Similar to FIRST or Fire Simulation Technique, CFAST recognise an enclosure as a structure made up to two zones with fix total volume, boundaries between layers that can move up and down depending on emission and ventilation process, and uniform condition in each zones . Figure 4 Example CFAST simulation result 2.2 Field Models Field models are based on the division of volume into large number of computational cells and application of finite difference technique. It requires fewer assumptions and can provide very complex solutions compared to zone models. As opposed to zone models, field models that are based on CFD or computational fluid dynamics can provided detailed spatial; information about the fire environment . Therefore, field models can represent an almost complete fire event and more detailed interaction between elements compared to zone models . Figure 5 – Computational Cells in Field Models Field models are often referred to as CFD, which is capable of predicting 3D distributions of velocity, fire products, and temperature. FDS or Fire Dynamics Simulator is a model for buoyancy-driven fluid flow from a fire . FDS combined zones and field model approaches and widely used fire model in fire protection engineering as it has not only capable of combustion modelling but Large Eddy Simulation or LES required in turbulence modelling. Moreover, FDS uses incompressible flow solver and perform Navier-Stokes equations . Figure 6 – FDS Simulation of the smoke flow from a kitchen fire FDS is compatible with large spaces such as atrium, auditorium, and others where zone models are not appropriate . However, since field models according to deal with large number of cells, it is computationally intensive and requires more memory, set-up time, extensive interpretation, sub grid models (i.e. turbulence models). 3 Application of Computer Modelling in Field Fire Investigation with Proof 3.1 Computer-based Fire Models for Fire Investigation The result of the literature survey conducted in 1992 suggests that there are 74 computer-based fire models in the world and they are divided in two general types of models – zone and field models . In terms of fire investigation however, only few models are suitable for this purpose. These include HAZARD I for tire hazard analysis, DETACT-T2 and DETACT-QS for predicting detector activation, ASET for egress time calculation, CCFM for multi-compartment zone fire mode, TENAB for tenability calculations, EXITT for exit time simulation, BREAK1 for glass breakage predictions, and FPETOOL for fire protection engineering calculations . Since the result of a particular fire is already known, fire investigation according to often concentrates on what was there before the fire and why they were found in a certain way after the fire. For this reason, investigators need an appropriate and effective fire model in order to understand the fire causes and fire development scenarios. The fire model is then validated according to the accuracy of their features particularly in determining heat release rate with values closer to a real fire . According to , computer fire modelling can help investigators understand the fire, provide important information about time, understand eyewitness accounts including fire progression in relation to other variables, and become more objective in analysing evidence. It can also help in survivability analysis such as how temperature, toxic gases, visibility, heat, and flame can affect people. Analyse post-fire indicators such as the effect of heat transfer on materials, visualisation of fire phenomena, test hypothesis and understand fire patterns as shown below, Figure 7 - Fire pattern analysis with Computer Fire Modelling (left) compared to Post-Fire Patterns (right) . 3.2 CFD Model Discovered the Cause of the 1987 Kings Cross Underground Fire One key aspect of 1987 King Cross Fire incident was the extremely rapid-fire development engulfing the entire ticket hall like a firestorm. Since no fire experts at that time can explain why the fire was so severe and rapid, a computer simulation of hot gases in the Kings Cross fire was carried out in 1992. A CFD Fire Model was used to simulate the flow of hot gases in the Piccadility line escalator tunnel to the ticket hall. The result of simulation as shown below drew attention to the indirect consequence of a combustion phenomenon now known as the “trench effect” . Figure 8 - CFD Model Grid layout of the ticket hall and the inflow and outflow boundaries employed in the simulation Figure 9 – Temperature distribution within the tunnel represented by flooded contours where lighter coloured regions are hot temperatures in the escalator trench. Figure 10 – Velocity vectors showing the presence of secondary flow along the escalator tunnel The cross-sectional view of the velocity vectors cutting through the heat source in Figure 8 (Right) demonstrating the trench effect. CFD revealed the convincing feature of fire spread in Kings Cross fire in the form of hot gases laid along the escalator, the combined effect of Conada and chimney effects, secondary flow along the tunnel, and preheated wooden escalator responsible for the unusually swift fire spread . 3.3 FDS and LS- DYNA for Simulating World Trade Centre Fire In 2005, a computer simulation of World Trade Centre fire was conducted in order to understand the collapse of the WTC towers. In order to determine the initial and boundary conditions for fire occurrence, LS-DYNA structural dynamics software provided the collision characteristic of the airplanes and the distribution of jet fuel including those consumed in the fireball. FDS computer code on the other hand modelled fire spread and determined the temperature and products of combustion from the WTC fire. Figure 11 - Upper Layer temperature contours of WTC 2, 81st floor The above figure demonstrate FDS potential to calculate the most likely upper layer temperature, distribution of the jet fuel spreading throughout the east side of the floor, fire migration and spread, and the contribution of window breaking times in increasing oxygen supply for fire . 3.4 FDS physical simulation of Station Nightclub Fire A fire occurred in the Station Nightclub located in West Warwick, Rhode Island in 2003. The fire killed one hundred people and the cause of the fire was pyrotechnics igniting foam insulation of the lining walls. The fire spread quickly along the ceiling and smoke was seen in exit doorways a minute after ignition . A physical simulation of the fire was conducted using FDS and the result was compared to images captured during the actual fire. FDS displayed the heat release rate and 3D smoke density, visibility, flame development after 300 seconds. Comparison between FDS and actual video footage was found reasonable. Moreover, simulation of tenability conditions inside the club suggests that open doors and windows provide sufficient fresh air to maintain a level of tenability which is consistent to video footage showing the last person exiting the building 250 seconds after ignition. Consequently, the findings of the investigation resulted to a number of improvements in codes, standards, and practices including installation of NFPA 13 compliant automatic fire sprinkler system in nightclubs, banning of non-fire-retarded polyurethane foam, NFPA 1126 standard on the use of pyrotechnics, increased factor of safety in occupancy limits calculation, and development of cost-effective computer fire models . 4 References Chan, C. (1996). Transport Phenomena in combustion, Taylor & Francis.UK Daeid, N. N. (2004). Fire Investigation, Taylor & Francis De Haan, J. & Icove, D. (2012). Kirk's Fire Investigation, Pearson Education.New Jersey FEMA (2013). Basic Tools and Resources for Fire Investigators: a Handbook, CreateSpace Independent Publishing Platform.USA Gann, R. & Friedman, R. (2014). Principles of Fire Behavior and Combustion, Jones & Bartlett Learning Gayev, Y. A., Hunt, J. C. R. & Division, N. A. T. O. P. D. (2007). Flow and Transport Processes with Complex Obstructions: Applications to Cities, Vegetative Canopies and Industry, Springer Gorbett, G. Computer Fire Models for Fire Investigation and Reconstruction. International Symposium on Fire Investigation Science and Technology, 2008 USA. ISFI, 1-12. Janssens, M. L. (2000). Introduction to Mathematical Fire Modeling, Second Edition, Taylor & Francis Karlsson, B. & Quintiere, J. (2002). Enclosure Fire Dynamics, Taylor & Francis Madrzykowski, D., Bryner, N. & Kerber, S. (2006). The NIST Station Nightclub Fire Investigation: Physical Simulation of the Fire, NIST.USA Maevski, I. Y., Board, N. R. C. T. R., Program, N. C. H. R., Highway, A. A. o. S., Officials, T. & Administration, U. S. F. H. (2011). Design Fires in Road Tunnels, Transportation Research Board National Research Council (1996). Fire- and Smoke-Resistant Interior Materials for Commercial Transport Aircraft, National Academies Press.USA NFPA (2008). Fire Protection Handbook, National Fire Protection Association.USA Rexfort, C. (2004). A Contribution to Fire Detection Modelling and Simulation, Tenea Wilkie, C. A. & Morgan, A. B. (2012). Fire Retardancy of Polymeric Materials, Second Edition, Taylor & Francis Yeoh, G. H. & Yuen, K. K. (2009). Computational Fluid Dynamics in Fire Engineering: Theory, Modelling and Practice, Elsevier Science Yung, D. (2008). Principles of Fire Risk Assessment in Buildings, Wiley  Read More