Modelo numérico y validación experimental para la simulación de incendios en edificios con recintos de gran volumen (atrios)

Abstract

In modern, industrialized and technologically developed societies, such as ours, there is a growing concern about risks and accidents prevention and safety. Fire safety represents one of the main aspects of this concern. The present thesis is focused on the study of fires within big volume interior enclosures buildings, which will be commonly named atrium from now on. This kind of structure has become a common element in modern architecture and can be found in high-rise buildings, auditoria, shopping centres, airports and mass transport stations, among others. The atrium represents an innovative, complex and non conventional architectonical element that can lead to fire environments diverging significantly from those used in the development of current codes and standards. Atria are a source of discussion in the fire safety community because smoke can easily spread from one floor to another, making the traditional methodologies for compartmentation of little or null effect. Moreover, detection, control and extinguishment of fires in atria differ significantly from those in small enclosures. With such remarkable architectural features, the fire dynamics in one building does not necessarily correspond to the fire dynamics in another building. Thus, a proper understanding of fire dynamics and smoke movement for each particular building is needed to provide the scientific understanding required in the proper design of fire-safe structures. Since the 80's, fires within atria started to be studied, both experimentally and numerically. However, testing in full-scale enclosures is too complex, expensive and labour intensive, resulting in a very small number of tests that can be reasonably carried out, and with reduced-scale testing, it is not easy to fully preserve the fluid, thermal and radiant similarities at the same time. Simple analytical and empirical correlations exist for smoke movement in atria but these only provide general behaviour and are usually not valid for complex designs like those in modern buildings. Another alternative is the use of advanced computer models. It is because of the subsequent improved understanding on fire dynamics and smoke management together with the increased computing power available nowadays and the improvements and developments of the existing numerical codes, that there is a current international trend in fire protection engineering regulations which is towards performance-based design and risk-informed analysis. Because of the shift towards performance-based codes and the difficulty of testing in atria, fire models are increasingly being used for developing fire safety engineering solutions. However, current development of performance-based and risk-informed fire protection needs more validation studies of fire modelling in atria. For this aim, it is essential to generate more reliable and comprehensive full-scale tests for the ongoing validation of fire models. Under the current situation, the two main objectives of the present thesis are to provide a broad set of new experimental data of atrium fires, as well as to check the capability of two numerical codes to simulate the fire environment induced in this kind of building, to help to fill the gap in experimental data and scientific understanding. In the present work, an overview of the current state-of-the-art of fires in atria is presented first. Next, a description of the experimental facility and the instrumentation used in the experiments is introduced. The fire tests conducted as part of the Murcia Atrium Fire Tests are presented. Later, the numerical models used to perform the experiments simulations are described. Finally, the most important results and numerical versus experimental comparisons are shown and discussed.