A comparison of a FLOW3D based fire field model with experimental room fire data

A Comparison Of Two Fire Field Models With Experimental Room Fire Data

“…For the FLOW3D results it was noted that while coarser meshes revealed similar overall flow features to those produced on the 8280 cell mesh, suggesting mesh independence, a significant mesh refinement to 45,356 cells (exploiting geometry symmetry) [12] reveals considerable improvements in the agreement with the measured results for some of the profiles (figure 4).…”

“…There appear to be no significant differences in the vertical velocity profile located in the centre of the door for the plate, single cell and two cell soffit representation [12]. However, the mass fluxes in and out of the compartment have changed from 0.500 kglsec (for both cases) as determined by the plate soffit to 0.551 kglsec (for both cases) for the two cell soffit, an increase of 10%.…”

“…A number of simulations were performed based on five fire locations, three door widths and two f i e sizes [12,13]. Here, the results for the 62.9 kW centre fire location will be reported.…”

Cfd Modelling Of Combustion And Heat Transfer In Compartment Fires

“…Although this might appear counter intuitive for a developed flow in a restricted passage, where wall boundary layer effects predominate, the flow in the doorway is not nearly so well-defined. Kumar et al [2], employing the JASMINE code, did not reproduce the measured trend, predicting higher velocities in the centre of the doorway, fig 5a) the experimentally observed trend towards h~gher velocities closest to the door jamb is clearly reproduced in the planes closest to the room ( RP and CP ) There is however some indication of a subsequent reversal of this trend In the plane hrthest from the room (AP) Such observations are broadly consistent with those reported in the earlier study [3] The predictions in the lower layer, on the other hand, do not exhibit such behaviour In general the computed velocities are lower adjacent to This situation is not substantially changed by fbrther mesh refinement. Whilst the velocity levels are then brought into closer agreement with experiment, consistent with the improved mass flux predictions in Table 1, the trend with increasing distance from the jamb remains unaffected.…”

“…The series of compartment fires investigated by Steckler et al [I] provides one of the more satisfactory test cases for field model prediction although the immediate vicinity of the fire source , where property changes are most pronounced, is also generally poorly resolved. A number of studies [2,3,4,5] have used this compartment fire as a field model test case but have largely concentrated on flowfield characteristics. In common with other similar applications, combustion is often entirely omitted [3,4,5] in favour of a simply prescribed heat source.…”