Thesis (M.S.) -- Worcester Polytechnic Institute. / Keywords: Hot smoke test; FDS; CFD; Computer modelling. Includes bibliographical references (p.105-110).

"The construction of smoke hazard management systems in large buildings such as shopping malls, cinemas, airports and train stations are increasingly being based on performance based design. Hot smoke tests are a method of using simulated fire conditions to evaluate the functionality of the completed building and the installed systems without causing damage. The author amongst others performed hot smoke tests (HST) according to Australian Standard AS 4391 -1999 in several buildings. In some tests air temperatures, air speed and smoke optical density were recorded at several locations during the test of which two tests are reported. These were later modelled by the author using Fire Dynamic Simulator (FDS) to show that typical fire protection engineering consultant applying the computer model may reasonably predict some results comparable to a full sized simulation scenario. However, some aspects were not well predicted. The modelling was improved by the outcomes of an investigation of the relationship between fuel properties, plume temperature and dynamics, and grid sensitivity. Areas of potential further improvement were identified. This work highlighted that the conditions witnessed in a hot smoke test can provide a guide, but do not represent all aspects of a real fire or design fire scenario. Although the FDS hot smoke model predicted comparable results to the hot smoke test, whilst suitable for system design, computer modelling should never be used as a system installation certification tool. Data from hot smoke tests, if gathered cost effectively, can be a valuable resource for computer model verification."