A "typical" hotel atrium with balconies, open to the atrium, which provide access/escape routes for the guest rooms was modeled with the FDS-1.0 for two different forced supply ventilation systems. One system supplied outside air from a fan room under the 5th floor ceiling, the other supplied outside air from slot vents at the top rear of the balconies. The total volume supply was the same for both systems. For the volume flow of balcony blowing used, the simulations showed no significant difference in the amount of smoke at head height on the balconies. One set of simulations placed the example fire on the atrium floor at a series of locations from under the lowest balcony to well away from the balcony edge. Smoke trapping on the balconies was most severe when the fire center was directly under the balcony edges. Another simulation placed the example fire in a 3rd level guest room. Smoke from this fire impinged on the balcony ceiling directly outside the guest room door and was deflected away from the balconies resulting in less smoke trapping for balconies immediately above than for comparable balconies for the atrium floor fire. A separate simulation modeled part of the atrium at a higher resolution than most of the simulations. Approximate boundary conditions were applied at the edges of the higher resolution simulation but always at least 2 meters away from the important parts of the flow being modeled. This simulation showed significantly less smoke at head height on the balconies. The balcony balustrades differed slightly between the two model resolutions, in particular they were thinner N-S in the higher resolution model. The episodic nature of the smoke flow onto the balconies suggests that the higher resolution results may better capture the local flow details and that the coarser grid results may be the more suspect. However, effects of door jets coming from the far end of the atrium were not modeled in the higher resolution, partial simulation, and may be equally important giving the lower resolution results some advantage. A full atrium simulation with an intermediate resolution in the vicinity of the fire showed an opposite trend - the intermediate resolution simulation showed more, not less smoke under the balconies than the base case. This tends to strengthen the suspicion that the boundary conditions applied to the high resolution case were inadequate. Physical experimental results could resolve this, but, with the computer and time available, higher resolution simulations of the full atrium could not be attempted.
展开▼