This paper describes an analytical study that was conducted to establish an accurate yet efficient algorithm for transporting fire data across disparate time scales between the fire and structural domains. In particular, a high-resolution computational fluid dynamics simulation was conducted using the computational fluid dynamics software Fire Dynamics Simulator (FDS) to measure the surface flux acting on an unprotected steel beam exposed to a heptane pool fire. Data from FDS was coupled to a heat transfer analysis in ABAQUS to determine the temperatures within the beam. Two subcycling algorithms were studied based on the conventional assumption of weak coupling (i.e., that the fire affected the structural response but the structural response did not affect the fire response). The first algorithm used a traditional subcycling technique in which surface fluxes from FDS were sampled at the requisite time steps in the thermal analysis. The second algorithm was formulated to produce a time-averaged flux at each time step in the thermal analysis. Results demonstrate that the time-averaged flux yields much greater accuracy than the time-sampled flux. Furthermore, the study shows that the time scale in the thermal analysis can be significantly relaxed if the energy equivalence of the surface flux is maintained.
展开▼
