| OCR Text |
Show im-i^rh d_ dx.' a,' dx,^ Mf Ctp-f where a,, C g and Cd are 0.7, 2.86 and 2.0, respectively. A gradient diffusion approximation is made to relate the transport of mixture fraction via turbulence fluctuations to mean mixture fraction gradients M, 3/ P*if = - - T - O. OX; Mean thermodynamic and chemical properties are then computed as Geometry/Mesh For these simulations, a 2-d axisymmetric grid consisting initially of 21,000 cells was generated using Fluent Inc.'s preprocessor, GAMBIT V1.0. The flexibility of unstructured meshing allows for regions of high gradients requiring a fine mesh to be sufficiently resolved while providing a coarser mesh to regions farther removed from the jet. Solution-based mesh adaption was utilized in all simulations to increase accuracy within the flame zone and ensure grid independence. The computational domain extends approximately 600 fuel nozzle diameters in each direction (See Figure 4). Though the nozzle configuration for this flame is trivially simple, flare flame nozzles can be quite complex, with geometry details significantly governing the ultimate Fuel andcoflow air inlets Burner wall Figure 4. Mesh for 2-d axisymmetric buoyant turbulent jet flame simulation. An unstructured quad mesh was used, allowing for efficient targeting of cell density. Solution-based grid adaption provides a highly effective means of local mesh refinement to improve accuracy and ensure grid independence. |
