Effects of visibility on the modeling of fire evacuation in a multi-storey passenger ship

• Main Author: Ho, Jasmine Bi-Yan.
• Other Authors: Tan Kang Hai
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Maritime studies::Maritime management and business
• Online Access: http://hdl.handle.net/10356/15916
• Institution: Nanyang Technological University

Analytical research on evacuation processes carried out in passenger ships has attracted increasing interests over the past few years, with increasing size of the ship. Most of the approaches utilized are flow models. In this paper, an extensive cellular automata (CA) model is proposed to simulate evacuations in a multi-storey passenger ship. CA has been successfully applied to simulate various complex systems, such as those of traffic and fluids. It also provides a more natural approach towards pedestrian dynamics. The effects of visibility will be studied with respect to the visibility range. To be more realistic, the proposed CA model also takes account of human behavior, tenability analysis and average movement velocity. A simple experiment was also conducted to find the average velocity of an individual while walking on flat ground and up a flight of stairs. These two actions were decided on as they are mainly involved in the movement towards the final holding area in a ship (i.e. open area at the highest deck), should there be an evacuation. The experimental details are attached in Appendixes A and B. In the numerical example, the phenomenon of “flow with the stream” is simulated. The results indicate that the different visibility range can affect an evacuation process significantly. It also shows that when the visibility range is wider, it will result in a higher evacuation rate. Thus by using the proposed CA model, a case study will be conducted. Currently, this CA model is used in emergency evacuation/fire safety assessments, such as drawing comparisons between different designs of evacuation schemes. However, it must be noted that the proposed CA model needs to be further improved. The results obtained by the proposed CA model cannot be validated, due to a lack of experimental information on human performance under emergency conditions. Therefore, real tests and experiments are extremely pivotal for future works. Lastly, the report closes with recommendations by the author which may be helpful in further research of related topics.