CFD Validation for Tunnel Smoke Control Design

With no reliable formula for determining minimum required upstream velocity to control smoke in tunnels, attention has turned to CFD for answers. However, CFD must be used with caution, as it produces many different answers, depending on the quality of the algorithms and coding, and how parameters and numerical models are selected. It is really only valid when the whole system of analysis (analyst + software + parameter/model choices) has been validated against known and relevant real fire scenarios. If the analyst is new, the software is different, solution parameters change, or there is no real case for comparison, perhaps it can no longer be relied on. A validation of a system of analysis against the best available tunnel fire data (Memorial Tunnel tests) is presented, with discussion on how different modelling options and software can affect the outcome, leading to conclusions as to how to model smoke control in tunnels that are not so different (in the physics and flow regimes present) to the Memorial Tunnel tests. How far the techniques might be stretched to different geometries or fire scenarios is also discussed briefly.