Transient Ceiling Jet Temperature and Velocity Profiles in the Presence of an Upper Layer: Comparison with Predictions by LAVENT and JET Computer Fire Models

Results of a scaled experimental study are compared with the predictions produced by the LAVENT and JET computer fire models. As in most large-scale fire testing, characterization of heat release rate (HRR) and accuracy of its measurement creates a lot of uncertainty for validation of these codes. Sensitivity of LAVENT results to the assumed radiative loss of 35% is examined by varying the total HRR which corresponds to varying the convective HRR. Comparisons show that LAVENT uses the wall-jet theory at its core for prediction of the ceiling jet flow and thus does not capture the ceiling jet profile well and overpredicts the upper layer temperature during the development of the layer. The velocity is generally overpredicted since the retardation of the jet momentum by the upper layer does not seem to be modeled accurately. JET has a better user interface and slightly different parameters, but does not provide ceiling jet profiles. The major difference between JET and LAVENT include the ceiling jet temperature and velocity algorithms, the fusible link algorithms, and the use of a variable radiative fraction as a function of fire size and type.