Multi-objective optimization of underground car park design for tenability under fire-induced smoke

The effects of the inclusion of design factors in optimization such as ceiling height, beam span length, transversal beam depth, longitudinal beam depth, and extraction fan rate on the tenability of car parks during fire were studied. Fire Dynamic Simulator (FDS) was employed as the simulation tool for response generation, from which the preliminary simulation results were firstly compared with published experimental data for validation. Responses such as smoke descent time, critical velocity, and temperature were optimized using the Response Surface Methodology (RSM). Thirty-two FDS models were constructed using the Central Composite Design (CCD) method and second-order models were formulated. The predicted responses from the second-order models were in good agreement with those obtained from FDS. In the optimized design, the smoke descent time was increased by 191.08% and the critical velocity was reduced by 61.67%. Meanwhile, the change of temperature was marginal. Moreover, by examining the flow result, the employment of two longitudinal beams from the optimized design could effectively channel the hot gasses to the downstream region. Overall, it was found that the presence of beams combined with the FDS statistical analysis could improve the tenability of enclosed car parks during fire