Loading...

Chemical kinetics Optimization for a DLE Combustor Optimization Using Numerical Method

Shakeri, Alireza | 2018

883 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 50713 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Mazaheri, Karim
  7. Abstract:
  8. Control and reduction of gas turbines pollutants emissions are increasingly of major concern. Here we introduce and apply new algorithms for derivation of accurate and low cost skeletal chemical kinetics, optimization of global multi-step chemical kinetics, and optimization of geometry of Lean Premixed (LPM) dry low emission (DLE) combustors.A new reduction algorithm called SA-GEIE, based on sensitivity analysis of reaction mechanisms respect to ignition delay, flame temperature and pollutants concentration (NO and CO), is introduced. This algorithm is implemented to GRI-3.0 to produce a low cost methane oxidation skeletal mechanism with 118 reactions and 39 specie. The accuracy is assessed for a range of equivalence ratios (0.5-2), pressures (1-30 bars) and initial temperatures (300-1800 K) using different reactor models e.g., PSR (partially stirred reactor), IGNITION, and PREMIX. The reduction algorithm is also effectively applied to a base mechanism for kerosene oxidation, and its applicability and accuracy is evaluated.Next, a new algorithm, called DORP (Differential-evolution Optimization of Rate Parameters), is introduced for optimization of parameters of global multi-step chemical kinetics. The algorithm uses one of CHEMKIN chemical reactor models, i.e., IGNITION, PSR or PSR-PFR models. The algorithm uses DE non-gradient optimization algorithm with a variety of optional cost functions (e.g., pollutant concentrations, ignition delay time, flame temperature, etc.). The algorithm is applied to a few five and eight steps global mechanisms for methane oxidations. For validation and assessment of applicability and accuracy, results are compared with GRI-3.0 and other existing models. The scheme is also applied to model oxidation of propane and ignition of kerosene by means of three step global mechanisms.To show applicability of DORP to real industrial applications, an optimized five-step global mechanism is used in a 2D CFD simulation to predict pollutant emissions of a known experimental reactor, called JSR. The simulation software uses RANS equations along with the species transport equations. RSM turbulence model and finite rate eddy dissipation model is used to simulate interaction of turbulence and combustion. Grid independence is shown, and hydrodynamic and thermal results, including concentration of pollutants, are compared with experimental data. Further, the simulation is used to analyze effect of residence time, equivalence ratio, inlet temperature, and combustor pressure on NO and CO production. It is shown that for LPM condition, reduction of inlet temperature and increase of pressure may reduce NO emissions.Finally, a DE optimization algorithm is integrated to a CFD solver, which uses our optimized low cost 5-step chemical kinetics to optimize geometry of the same JSR combustor for the least possible NO emissions. The geometry of combustor is parameterized using three parameters, and these parameters are optimized, to reduce NO emissions more than 10%
  9. Keywords:
  10. Differential Evolution Algorithm ; Chemical Kinetic ; Combustion Chamber ; DE Optimizer ; SA-GEIE Algorithm ; Multi-Step Global Mechanisms ; Differential-Evolution Optimization of Rate Parameters (DORP)Algorithm ; JSR Combustor

 Digital Object List