Effects Of Computational Domain On Field Modeling Of Building Fires

As an important component of numerical simulation, the selection of computational domain has begun to attract more and more attention of scholars both at home and abroad. However, systematic studies on the selection of computational domain during a numerical simulation are still relatively rare, of which, one reason can be attributed to the lack of method for quantitatively analyzing the rationality of computational domain selection, and the other can be attributed to the restriction of many affecting factors and the difficulties to quantify the computational domain selection.Centring on the issues related to effects of computational domain on field modeling of building fires and taking a general construction as an example, this article, firstly, carried out severl types of numerical simulations, mainly involving the effects of computational domain on the FDS simulations of two major fires under three different shapes of building opening. Secondly, with a shape of building opening selected as the subject, the changes of computational domain selection due to the fire types were extensively investigated. The main contents included:1. With the grid and computational domain crossly examined, the FDS results of a fire power were compared with corresponding experimental results. Then, combined with the correlation analysis method, a reasonable and economic grid size sutible for such a fire release rate was finalized. Further, according to related studies, a general method for grid selection was proposed.2. Combined with the correlation analysis method, series of groups of numerical simulations were carried out, involving the effects of computational domain on the FDS simulations of two major fires under three different shapes of building opening, which indicated the variations of simulation results (including temperature and velocity) with growing computational domain extension.3. Based on the preceding analysis, by comparing the simulation results of three shapes of building opening, the internal relation between the computational domain extension and the shape of building opening as well as the fire types was revealed.4. By the dimensionless analysis method, two dimensionless quantities, i.e. the dimensionless computational domain extension and the dimensionless fire power, were constructed. And on the basis of the acquired data, a statistical analysis was carried out about the two dimensionless quantities, of which, the result revealed the variation of dimensionless fire power with the dimensionless computational domain extension. Finally, based on the above and proceeding from the practical application of fire engineering, a dimensionless model of computational domain extension for the FDS simulation of building fires was proposed.