Fire engineers commonly use computational fluid dynamics (CFD)-based fire modelling to develop and test the designs of complex buildings. However, as these models are very complicated and this type of simulation takes an impractically long time to run, engineers typically reduce the size of the model to shorten the time needed.
The problem is that a simplified model may not be able to account for the entire building. That means that the ways in which fire and smoke might spread throughout the whole building cannot be fully determined, nor can the effect on the growth and spread of fire and smoke of parts of the building far from the fire. This can put limits on the ability of engineers to develop robust fire safety solutions.
One way of addressing this limitation is the use of ‘coupled hybrid modelling’. This combines the originally used model (which takes a long time to run) and another model (which takes a much shorter time) together. Fire engineers can thereby increase the model size to include more or all of a building, but still carry out a simulation in a reasonable length of time.
A literature review of existing coupled hybrid models identified a lack of collaborative working and the use of proprietary software. Therefore, a coupled hybrid model has been developed, based on the open-source and widely used fire model, Fire Dynamics Simulator (FDS).
A new computer model must be compared with real experiments to confirm that it can represent the real world within reasonable limits. Applicable experimental data were not found during an extensive literature search and discussions with applicable bodies. Therefore, to provide these data, a new experimental rig was specifically designed and built.
Data from the new rig were first used to examine potential holes in the current typical fire safety design model for complex buildings. This analysis showed that there are phenomena related to the two-way coupling of a fire and a shared ventilation system, which may not be captured by commonly used simple engineering methods – or addressed in current design guidance documents.
The new experimental data was then compared with the predictions of the coupled hybrid computer model, and the qualitative and quantitative differences highlighted. The computer model was found to predict fire hazards within the experimental set-up effectively, providing key information needed by future model users to ensure that their analysis is robust.
The original insights into how a total building system and a fire interacts, and the newly developed coupled hybrid model, can be used together by fire engineers. This will help them to deliver design solutions with more deeply understood fire life safety risks for complex buildings that incorporate mechanical ventilation systems.
The overall experience and opportunity has been priceless and I have learnt so much. I now aim to continue expanding my knowledge and learning new skills to help society become safer and more sustainable.
-Benjamin Ralph speaking about his BRE Trust supported PhD at the University of Edinburgh.