Study of the sensitivity of Fire Dynamics Simulator and its applicability in different fields of fire research

Abstract

The evolution of computer technology has allowed for the development of computer simulation tools in the field of fluid dynamics (CFD), enabling the exploration of various aspects of engineering with a precision in calculation that would otherwise be unattainable. Fire engineering, or fire protection engineering, involves analyzing the behavior of fires and designing solutions intended to be applied to buildings and industries to protect people and structures from the effects of fire. The use of CFD tools is well established in engineering fields such as chemical, mechanical, or aeronautical engineering. However, in fire engineering, the complexity of the accidental fire phenomenon, as well as the degree of uncertainty it presents, makes its application more challenging and the acceptance of justifications based on simulations is lower than it would be in other fields. The Fire Dynamics Simulator (FDS), developed by the National Institute of Standards and Technology (NIST) of the United States, is the reference software for fire simulation in the field of fire engineering. It is commonly used to justify design solutions for fire protection that differ from those specified in fire regulations. It has also been used forensically for fire reconstruction as well as in various areas of fire science research. This work was intended to study the various phenomena that explain the behavior of compartment fires and to relate the physical and chemical principles governing them to the models and simplifications used by FDS software in order to evaluate the capabilities and limitations of this software. Simulations were conducted in different scenarios to study the program's ability to reproduce some phenomena of interest, as well as to analyze the sensitivity of the results to design decisions in characterizing input variables. FDS presented strong capabilities in representing fire phenomena related to the movement of smoke and hot gases, even in rapidly changing scenarios. However, when input parameters such as grid size or fire location were slightly changed, significant variations in output emerged, which could mislead the interpretation of the results