Search

Computational fluid dynamics (CFD) modeling (or field modeling) is becoming the main method for numerical simulation of building fires. Among many factors that influence the validity and accuracy of CFD simulation results, the computational domain is sometimes overlooked. In this article, the effects of computational domain on simulation results are analyzed. Simulation results from the use of different domains are compared with experimental data reported in the literature. A parametric study is then conducted to reveal a relationship between the effective domain extension and the heat release rate of the enclosure fire. The effect of computational domain extension in relation to vent opening is also investigated. It is found that the selection of computational domain can have a significant effect on the outcome of enclosure fire simulations. Determination of the appropriate computational domain without unduly sacrificing computational efficiency is also discussed. [ABSTRACT FROM PUBLISHER]

In response to the call for the development of risk informed, performance-based building fire regulations, a literature review was conducted to link the traditional safety factor approach in fire safety engineering to failure probability. It can be demonstrated that a safety factor alone is an insufficient measure, or is only a first-order measure of risk or failure probability. To achieve a higher order estimate of failure probability, an α-percentile method is proposed. As part of this methodology, it is also proposed that the fire engineering briefing process involve nominations of percentile values for design fire scenarios as well as other parameters that define the scenarios. [ABSTRACT FROM AUTHOR]

The influence of multiple parameters on smoke exhaust efficiency in an atrium is the subject of this article. Detailed information is obtained through a series of four numerical simulations, all but one case involving a mechanical exhaust system. Quantitative smoke exhaust efficiencies in various cases are evaluated and compared. The investigated parameters include the activation time of the smoke exhaust system, the mechanical fan capacity, the elevation of make-up air inlets and the horizontal location or distributed arrangement of the inlet distribution. It is revealed that all these parameters have impacts on smoke exhaust efficiency to different degrees and hence on the performance of the smoke exhaust system. Improper design and operation may lead to a significant reduction in efficiency. The mechanism that causes the reduction of efficiency is discussed. Velocity and temperature fields are analyzed to reveal the interactions between the fire plume and the inlet jet flows. [ABSTRACT FROM AUTHOR]

This article presents a risk, or probability-based method for determining the residual fire resistance that is required when sprinklers are installed supplementary to building regulatory requirements. The concept of equivalence of a building design to a code-complying design on the basis of risk is introduced. The probability of failure is formulated, taking into account the probability density distributions for fire resistance of building structures and for fire severity. The sprinkler system is treated as a dual-exclusive-status system which has a bi-valued probability distribution. The probability density distribution for fire severity is moderated by the discrete reliability for sprinklers. A parametric study is conducted to demonstrate the sensitivity of the residual fire resistance to sprinkler reliability and fire severity distribution. [ABSTRACT FROM AUTHOR]

Fire modeling is an essential part of risk-cost assessment. Fire models, particularly those of zone type, have flourished in recent years. The selection of a fire model as a sub- model for a fire safety system model needs further verification and review in a number of aspects. In the present study, the performances of two fire models, namely NRCC and CFAST, are evaluated with respect to accuracy, efficiency and simplicity. This study indicates that discrepancies exist between the experimental results and the predictions of both models for one type of fire scenario or another. The merits and weaknesses of the two models are identified in regard to the demands of a risk assessment system model. [ABSTRACT FROM PUBLISHER]