guest ::
| Home > Online First > Improvement and Validation of Visibility Models in Fire Safety |
| Home > Online First > Improvement and Validation of Visibility Models in Fire Safety |
| Book/Dissertation / PhD Thesis | FZJ-2025-05139 |
2025
Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag
Jülich
ISBN: 978-3-95806-857-5
This record in other databases:
Abstract: In case of fire, smoke poses the major threat to the occupants of public or residential buildings. Notably, besides the smoke’s toxicity, reduced visibility can prevent people from safe egress. Commonly applied fire models to assess visibility at the building’s design stage often rely on input quantities that could not be consistently validated across multiple studies. Furthermore, the common interpretation of simulation results does not accurately mirror real-world conditions. This dissertation aims to develop robust methods for the experimental validation of visibility models and the enhanced post-processing of simulation results in the context of performance based design. The thesis comprises three publications exploring these topics. The first publication introduces the improvement of an existing photometric method (LEDSA) to measure temporally and spatially resolved extinction coe!cients in laboratory test fires. LEDSA involves capturing the smoke induced change in intensity of LEDs using consumer digital cameras. Validating against the established MIREX measurement system revealed, that utilizing RAW image data significantly improves measurement accuracy compared to the previously used JPG files. While using higher quality LEDs helped to increase reproducibility of the measurements, further uncertainties of the model and the experimental setup could be identified. In the second publication, LEDSA is compared on multiple test fires to the Radiance Method, which was developed at the University of Waterloo. It involves measuring the contrast on adjacent light and dark areas in image or video footage. The Radiance Method measurements align well with LEDSA and the MIREX for smoke from n-heptane pool fires, with much lower computational e"ort than LEDSA. However, only LEDSA was in line with the MIREX measurements for wood smouldering fires, while the Radiance Method could only reproduce patterns of the reference measurement. The third publication shifts focus from model validation to application, introducing visibility maps as a novel approach to assess visibility in performance based design. By post-processing existing data from fire simulations, the maps indicate areas where exit signs remain visible along the route of egress. Integrating the extinction coe!cient along the line of sight allows applying Jin’s empirical correlation to non-homogeneous smoke environments, providing for a more realistic assessment of visibility, than the classical way of treating it as a local quantity. Visibility maps can also account for additional factors like view angle and visual obstruction of exit signs.
|
The record appears in these collections: |
PDF