OCR Text |
Show YTTCL. Numerical Modeling of Turbulent Jet Diffusion Flames: Challenges and Implications for Flare Flame Simulations Eric S. Bish, Ph.D. Thomas S. Norton, Ph.D. Fluent, Inc. 10 Cavendish Court Centerra Resource Park Lebanon, NH 03766-1442 USA Flare systems play an important role in the petrochemical, chemical processing, mining and waste handling industries. These systems need to perform satisfactorily under a wide range of operating conditions. Computational Fluid Dynamics (CFD) can be a powerful tool in conjunction with experimental testing in evaluating flare system performance to ensure reliability and safety. Utilizing C F D fully in the design of flare systems requires an understanding of the relevant physics as well as the current modeling tools available. To gain further insight into the numerical modeling of flare flames, C F D results using the standard k-e, realizable k-e and Reynolds stress turbulence models in F L U E N T V5 were compared with experimental measurements in a buoyant, turbulent jet diffusion flame. Over the range of data available, all models perform comparably. Good agreement is seen for mean profiles of velocity, mixture fraction and temperature. Experimental data and simulation results compare well with established scaling laws for jet spreading rate and centerline momentum flux density. Based on these results, the realizable k-e model provides the most accurate, cost-effective solution for turbulence modeling in flare flames. Further investigation of these models in the buoyancy dominated, plume regime is suggested. Introduction Flare systems can be found in a wide variety of industries, including off-shore oil and gas drilling operations, refineries, processing plants, mines and landfills. Gas or liquid fuel may be burned, with the jet(s) issuing in an arbitrary direction (e.g., Figures 1,2). They need to perform reliably and safely and burn cleanly under a wide range of fuel flow and ambient conditions. Because of their typically large scale and broad range of anticipated operating conditions, comprehensive, full scale testing of flare systems is prohibitively difficult and costly. Considering these obstacles, Computational Fluid Dynamics (CFD) provides a powerful tool for investigating flare systems. Detailed parametric studies can provide critical insights into optimal design configurations while dramatically reducing design cycle times and costs. But to utilize CFD effectively in the design of flare systems requires an understanding of the relevant physics and the current modeling tools available. Flares generally consist of one or more turbulent jet diffusion flames. This is a seemingly simple configuration, but many complexities exist. These include the wide range of physical length and time scales inherent in the turbulent flow, the critical role of turbulence in determining the jet entrainment rate, turbulence-chemistry interactions and the influence of buoyancy on the flow of hot combustion products. Accurate modeling of these phenomena, and in particular flow turbulence and mixing, is important in flare flames where air entrainment governs significantly the mixing, chemical reactions and radiative intensity of the flame. In this paper, CFD simulations of a turbulent jet diffusion flame using the standard k-e, realizable k-e and Reynolds stress turbulence models in FLUENT V5 are For presentation at the 1998 American/Japanese Flame Research Committee Int'l Symposium, Maui, Hawaii. October 11-15, 1998. |