Integration of a 1D model with FDS for multiscale analysis of tunnels

Numerical simulations are used to reduce the number of tests required in a lot of scientific fields. It works in that way in the field of Fire science with the usage of CFD (Computational Fire Dynamics). Fire simulations take less time to complete as computer sciences advance. But tunnel simulations with long domains still take long times limiting the opportunities to develop applications in fields that require live time results, like risk assessment, emergency systems, etc. A Multiscale algorithm is presented. This algorithm integrates Whitesmoke, a 1D algorithm developed to simulate fluid flow in networks, into FDS (Fire Dynamics Simulator), a 3D LES program used to simulate fire dynamics. The aim of this integration is optimizing both the calculation time and accuracy, using the fast solutions of the 1D in uniform zones and the detailed solutions of the FDS in complex areas. The accuracy of the Multiscale is evaluated by comparing it to full 3D simulations. In this case, a tunnel of 4.8m x 4.8m and 600m of length is simulated. The flow velocities and temperature of Multiscale and FDS simulations are compared. The Multiscale model achieves a time saving that is closely proportional to the portion of the domain calculated with the 1D sub-model. And, even when the simulation time is shorter the difference with the outputs obtained by the FDS is small in temperature, velocities and backlayering extension. The presented model is capable of reducing the time necessary to make a tunnel fire simulation without jeopardizing its results. Still, the Multiscale has some areas to improve and develop, as its boundary conditions, which should be improved further in the future.