The Optimisation of Reaction Rate Parameters for Chemical Kinetic Modelling Using Genetic Algorithms

2002 ◽  
Author(s):  
L. Elliott ◽  
D. B. Ingham ◽  
A. G. Kyne ◽  
N. S. Mera ◽  
M. Pourkashanian ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Gas Turbine ◽  
Rate Constants ◽  
Reaction Rate ◽  
Kinetic Modelling ◽  
Burning Velocity ◽  
Chemical Kinetic ◽  
Rate Coefficients ◽  
Inversion Process ◽  
Experimental Findings

It is well recognised that many important combustion phenomena are kinetically controlled. Whether it be the burning velocity of a premixed flame, the formation of pollutants in an exhaust stack or the conversion of NO to NO2 in a gas turbine combustor, it is important that a detailed chemical kinetic approach be undertaken in order to fully understand the chemical processes taking place. This study uses a genetic algorithm to determine new reaction rate parameters (A’s, β’s and Ea’s in the Arrhenius expressions) for the combustion of both a hydrogen/air and methane/air mixture in a perfectly stirred reactor. In both cases, output species profiles obtained from an original set of rate constants are reproduced by a new different set obtained using a genetic algorithm inversion process. The new set of rate constants lie between predefined boundaries (±25% of the original values) which in future work can be extended to represent the uncertainty associated with experimental findings. In addition, this powerful technique may be used in developing reaction mechanisms whose newly optimised rate constants reproduce all the experimental data available, enabling a greater confidence in their predictive capabilities. The results of this study therefore demonstrate that the genetic algorithm inversion process promises the ability to assess combustion behaviour for fuels where the reaction rate coefficients are not known with any confidence and, subsequently, accurately predict emission characteristics, stable species concentrations and flame characterisation. Such predictive capabilities will be of paramount importance within the gas turbine industry.

A Novel Approach to the Optimisation of Reaction Rate Parameters for Methane Combustion Using Multi-Objective Genetic Algorithms

2003 ◽  
Author(s):  
L. Elliott ◽  
D. B. Ingham ◽  
A. G. Kyne ◽  
N. S. Mera ◽  
M. Pourkashanian ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Reaction Mechanism ◽  
Reaction Rate ◽  
Flame Structure ◽  
Simulated Data ◽  
Methane Combustion ◽  
Rate Coefficients ◽  
Inversion Process ◽  
Multi Objective ◽  
Novel Approach

This study uses a multi-objective genetic algorithm to determine new reaction rate parameters (A’s, β’s and Ea’s in the non-Arrhenius expressions) for the combustion of a methane/air mixture. The multi-objective structure of the genetic algorithm employed allows for the incorporation of both perfectly stirred reactor and laminar premixed flames data in the inversion process, thus enabling a greater confidence in the predictive capabilities of the reaction mechanisms obtained. Various inversion procedures based on reduced sets of data are investigated and tested on methane/air combustion in order to generate efficient inversion schemes for future investigations concerning complex hydrocarbon fuels. The inversion algorithms developed are first tested on numerically simulated data. In addition, the increased flexibility offered by this novel multi-objective GA has now, for the first time, allowed experimental data to be incorporated into our reaction mechanism development. A GA optimised methane air reaction mechanism is presented which offers a remarkable improvement over a previously validated starting mechanism in modelling the flame structure in a stoichiometric methane-air premixed flame (http://www.leeds.ac.uk/ERRI/research/res.html). In addition, the mechanism outperforms the predictions of more detailed schemes and is still capable of modelling combustion phenomena that were not part of the optimisation process. Therefore, the results of this study demonstrate that the genetic algorithm inversion process promises the ability to assess combustion behaviour for fuels where the reaction rate coefficients are not known with any confidence and, subsequently, accurately predict emission characteristics, stable species concentrations and flame characterisation. Such predictive capabilities will be of paramount importance within the gas turbine industry.

A Novel Approach to the Optimization of Reaction Rate Parameters for Methane Combustion Using Multi-Objective Genetic Algorithms

2004 ◽  
Vol 126 (3) ◽  
pp. 455-464 ◽  
Author(s):  
L. Elliott ◽  
D. B. Ingham ◽  
A. G. Kyne ◽  
N. S. Mera ◽  
M. Pourkashanian ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Reaction Mechanism ◽  
Reaction Rate ◽  
Flame Structure ◽  
Simulated Data ◽  
Methane Combustion ◽  
Premixed Flame ◽  
Rate Coefficients ◽  
Inversion Process ◽  
Multi Objective

This study uses a multi-objective genetic algorithm to determine new reaction rate parameters (A’s, β’s and Ea’s in the non-Arrhenius expressions) for the combustion of a methane/air mixture. The multi-objective structure of the genetic algorithm employed allows for the incorporation of both perfectly stirred reactor and laminar premixed flame data in the inversion process, thus enabling a greater confidence in the predictive capabilities of the reaction mechanisms obtained. Various inversion procedures based on reduced sets of data are investigated and tested on methane/air combustion in order to generate efficient inversion schemes for future investigations concerning complex hydrocarbon fuels. The inversion algorithms developed are first tested on numerically simulated data. In addition, the increased flexibility offered by this novel multi-objective GA has now, for the first time, allowed experimental data to be incorporated into our reaction mechanism development. A GA optimized methane-air reaction mechanism is presented which offers a remarkable improvement over a previously validated starting mechanism in modeling the flame structure in a stoichiometric methane-air premixed flame (http://www.personal.leeds.ac.uk/∼fuensm/project/mech.html). In addition, the mechanism outperforms the predictions of more detailed schemes and is still capable of modeling combustion phenomena that were not part of the optimization process. Therefore, the results of this study demonstrate that the genetic algorithm inversion process promises the ability to assess combustion behavior for fuels where the reaction rate coefficients are not known with any confidence and, subsequently, accurately predict emission characteristics, stable species concentrations and flame characterization. Such predictive capabilities will be of paramount importance within the gas turbine industry.

Review of Hybrid Emissions Prediction Tools and Uncertainty Quantification Methods for Gas Turbine Combustion Systems

2017 ◽  
Author(s):  
Sajjad Yousefian ◽  
Gilles Bourque ◽  
Rory F. D. Monaghan
Keyword(s):  
Uncertainty Quantification ◽  
Gas Turbine ◽  
Kinetic Modelling ◽  
Computational Cost ◽  
Chemical Kinetic ◽  
Design Development ◽  
Deterministic Models ◽  
Prediction Tools ◽  
Gas Turbine Combustion ◽  
Combustion Systems

There is a need for fast and reliable emissions prediction tools in the design, development and performance analysis of gas turbine combustion systems to predict emissions such as NOx, CO. Hybrid emissions prediction tools are defined as modelling approaches that (1) use computational fluid dynamics (CFD) or component modelling methods to generate flow field information, and (2) integrate them with detailed chemical kinetic modelling of emissions using chemical reactor network (CRN) techniques. This paper presents a review and comparison of hybrid emissions prediction tools and uncertainty quantification (UQ) methods for gas turbine combustion systems. In the first part of this study, CRN solvers are compared on the bases of some selected attributes which facilitate flexibility of network modelling, implementation of large chemical kinetic mechanisms and automatic construction of CRN. The second part of this study deals with UQ, which is becoming an important aspect of the development and use of computational tools in gas turbine combustion chamber design and analysis. Therefore, the use of UQ technique as part of the generalized modelling approach is important to develop a UQ-enabled hybrid emissions prediction tool. UQ techniques are compared on the bases of the number of evaluations and corresponding computational cost to achieve desired accuracy levels and their ability to treat deterministic models for emissions prediction as black boxes that do not require modifications. Recommendations for the development of UQ-enabled emissions prediction tools are made.

Effect of the potential-energy surface on the diatom dissociation rate constant for F 2 in Ar at 3000 K

1987 ◽  
Vol 414 (1847) ◽  
pp. 297-314
Keyword(s):  
Steady State ◽  
Rate Constant ◽  
Rate Constants ◽  
Reaction Rate ◽  
Energy Surface ◽  
Rate Coefficient ◽  
Classical Mechanics ◽  
Rate Coefficients ◽  
Steady State Rate ◽  
State Rate

Gas-phase dissociation of fluorine ( 1 Ʃ + g ) molecules in an agron bath at 3000 K was studied by using the 3D Monte Carlo classical trajectory (3DMCCT) method. To assess the importance of the potential energy surface (PES) in such calculations, three surfaces, with a fixed, experimentally determined F 2 dissociation energy, were constructed. These surfaces span the existing experimental uncertainties in the shape of the F 2 potential. The first potential was the widest and softest; in the second potential the anharmonicity was minimized. The intermediate potential was constructed to ‘localize’ anharmonicity in the energy range in which the collisions are most reactive. The remaining parameters for each PES were estimated from the best available data on interatomic potentials. By using the single uniform ensemble (SUE) method (Kutz, H. D. & Burns, G. J. chem. Phys . 72, 3652-3657 (1980)), large ensembles of trajectories (LET) were generated for the PES. Two such ensembles consisted of 30000 trajectories each and the third of 26200. It was found that the computed one-way-flux equilibrium rate coefficients (Widom, B. Science 148, 1555-1560 (1965)) depend in a systematic way upon the anharmonicity of the potential, with the most anharmonic potential yielding the largest rate coefficient. Steady-state reaction-rate constants, which correspond to experimentally observable rate constants, were calculated by the SUE method. It was determined that this method yields (for a given trajectory ensemble, PES and translational temperature) a unique steady-state rate constant, independent of the initial, arbitrarily chosen, state (Tolman, R. C. The principles of statistical mechanics , p. 17. Oxford University Press (1938)) of the LET, and consequently independent of the corresponding initial value of the reaction rate coefficient. For each initial state of the LET, the development of the steady-state rate constant from the equilibrium rate coefficient was smooth, monotonic, and consistent with the detailed properties of the PES. It was found that, although the increased anharmonicity of the F 2 potential enhanced the equilibrium rate coefficients, it also enhanced the non-equilibrium effects. As a result, the steady-state rate constants were found to be insensitive to the variation of the PES. Thus, the differences among the steady-state rate constants for the three potentials were of the order of their standard errors, which was about 15% or less. On the other hand, the calculated rate constants exceeded the experimental rate constant by a factor of five to six. Because within the limitations of classical mechanics the calculations were ab initio , it was tentatively concluded that the discrepancy of five to six is due to the use of classical mechanics rather than details of the PES structure.

Simulation of Crosslink and Graft Mechanism of Hydrogenated-NBR/ZINC Di-Methacrylate Composites by Moving Die Rheometer

2002 ◽  
Vol 75 (5) ◽  
pp. 943-954 ◽  
Author(s):  
Takashi Sato ◽  
Akio Fujino ◽  
Sachio Hayashi ◽  
Motofumi Oyama ◽  
Katsumichi Ono
Keyword(s):  
Experimental Data ◽  
Genetic Algorithm ◽  
Rate Constants ◽  
Reaction Rate ◽  
Addition Reaction ◽  
Algorithm Optimization ◽  
Reaction Rate Constants ◽  
Data Comparison ◽  
Graft Reaction ◽  
Comparison With Experimental Data

Abstract This paper studies the crosslink and graft mechanism of hydrogenated-NBR/zinc di-methacrylate (HNBR/ZDMA) composites. The experiments to determine the rate constants of crosslink and graft reaction formulas are from Moving Die Rheometer (MDR) data. Comparison with experimental data and the solution of simultaneous ordinary differential equations for the crosslink and graft reactions are performed. Genetic algorithm optimization of the reaction rate constants allowed for simulation of crosslinking and graft reaction process. This simulation predicts that the HNBR/peroxide cure system will contain 0.1052-mole/l crosslink between matrix HNBR polymer at 2 phr.-peroxide concentration. The crosslink concentration of 0.1052-mole/l is composed of 0.0308-mole/l-addition reaction and 0.0744-mole/l-termination reaction. Moreover, the HNBR/ZDMA cure system contains 0.0148-mole/l graft of poly-ZDMA to HNBR and 0.0903-mole/l crosslink between matrix HNBR polymer at a 2 phr.-peroxide concentration.

scholarly journals Technical note: Monte Carlo genetic algorithm (MCGA) for model analysis of multiphase chemical kinetics to determine transport and reaction rate coefficients using multiple experimental data sets

2017 ◽  
Vol 17 (12) ◽  
pp. 8021-8029 ◽  
Author(s):  
Thomas Berkemeier ◽  
Markus Ammann ◽  
Ulrich K. Krieger ◽  
Thomas Peter ◽  
Peter Spichtinger ◽  
...  
Keyword(s):  
Experimental Data ◽  
Genetic Algorithm ◽  
Monte Carlo ◽  
Kinetic Models ◽  
Reaction Rate ◽  
Rate Coefficients ◽  
Model Parameters ◽  
Data Sets ◽  
Model Input ◽  
Input Parameters

Abstract. We present a Monte Carlo genetic algorithm (MCGA) for efficient, automated, and unbiased global optimization of model input parameters by simultaneous fitting to multiple experimental data sets. The algorithm was developed to address the inverse modelling problems associated with fitting large sets of model input parameters encountered in state-of-the-art kinetic models for heterogeneous and multiphase atmospheric chemistry. The MCGA approach utilizes a sequence of optimization methods to find and characterize the solution of an optimization problem. It addresses an issue inherent to complex models whose extensive input parameter sets may not be uniquely determined from limited input data. Such ambiguity in the derived parameter values can be reliably detected using this new set of tools, allowing users to design experiments that should be particularly useful for constraining model parameters. We show that the MCGA has been used successfully to constrain parameters such as chemical reaction rate coefficients, diffusion coefficients, and Henry's law solubility coefficients in kinetic models of gas uptake and chemical transformation of aerosol particles as well as multiphase chemistry at the atmosphere–biosphere interface. While this study focuses on the processes outlined above, the MCGA approach should be portable to any numerical process model with similar computational expense and extent of the fitting parameter space.

Sign in / Sign up

Export Citation Format

Share Document