Method of Kinetic Model Reduction for Computational Fluid Dynamics Applications

Leylegian, John C.; Paul, Tamal V.; Tulino, Vincent A.

doi:10.2514/1.b34805

Cited by 4 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As in Ref. 16, the model reduction process begins by using the constant-pressure streamtube and laminar flame speed to characterize the overall behavior of the model. The constant-pressure streamtube response is quantified in terms of the time necessary to achieve 50% of the temperature rise, rather than the four points of the temperature-time curve used in Reference 16.…”

Section: Methodsmentioning

confidence: 99%

“…The interaction coefficients were calculated at each time in the stoichiometric constant-pressure streamtube calculation; the change in temperature and species over the course of the reaction can significantly change influence coefficients, as seen in previous work. 16 Therefore as in Ref. 16, the influence coefficients were temperatureaveraged: The relative importance of each species was then determined using the temperature averaged influence coefficients, using a method similar to that used in the previous study.…”

Section: A Skeletal Model Generationmentioning

confidence: 99%

“…16, the influence coefficients were temperatureaveraged: The relative importance of each species was then determined using the temperature averaged influence coefficients, using a method similar to that used in the previous study. 16 All of the values of r ABT are sorted in descending order and interrogated. If for a given combination of A and B, a value of ρ has been assigned to A, but not to B, the value of ρ B is set to r ABT .…”

Section: A Skeletal Model Generationmentioning

confidence: 99%

“…In the previous study 16 the DRGEP method was used. For a kinetic model consisting of I reactions in J species, the rate of reaction of the i th reaction is:…”

Section: A Skeletal Model Generationmentioning

confidence: 99%

“…CHEMKIN. 15 A method was developed 16 using a combination of the directed relation graph method with error propagation (DRGEP) 13 and the solution mapping method 17 to reduce a model and to optimize it. The objective of the work presented here is to present improvements made to the method.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Incorporation of Path Flux and Steepest Descent Methods in Kinetic Model Reduction for CFD Applications

Abdulrahman¹,

Leylegian²

2014

50th AIAA/ASME/SAE/ASEE Joint Propulsion Conference

Self Cite

View full text Add to dashboard Cite

This paper outlines updates made to a method for chemical kinetic model reduction. The original method used the directed relation graph with error propagation to generate a skeletal model, and the solution mapping method to tune the remaining rate constants to match reduced model responses to those of the original detailed model. In the current work, skeletal model generation is performed using a path flux analysis, and changes have been made to the optimization process, by including an intermediate steepest-descent method in order to make the optimization process more efficient. The new method is then applied to the reduction of a methane-air model for two different sets of initial conditions. Results show rapid optimization of the model is possible as compared to previous work. Results for the optimized models also are able to predict the perfectly-stirred reactor response better than the optimized models generated using the older method. Nomenclature A i= preexponential constant for i th forward reaction rate constant b 0 = zero-order response function fit coefficient b p = first-order response function fit coefficients b p,q = second-order response function fit coefficients b p,q,s = third-order response function fit coefficients C A = consumption rate of species A C AB = consumption rate of species A to form species B E i = activation energy for i th forward reaction rate constant F i = correction term for i th forward reaction rate constant based on third body efficiencies, etc. f p = half-span of variation for the p th active parameter f = set of functions describing state of chemically reacting system g = set of differential equations describing time evolution of chemically reacting system I = number of reactions i = reaction index J = number of species j = species index k fi , k bi = forward and backward rate constants for i th reaction M = number of targets for optimization m = target index n = summation index P = number of active parameters for optimization P A = production rate of species A P AB = production rate of species A from species B p, q, s = active parameter indices R = universal gas constant r AB = interaction between species A and species B r ABT = temperature-averaged value of r AB S = sensitivity coefficient = laminar flame speed T = temperature t = reaction time t 50% = reaction time to achieve 50% of total temperature rise X p = factorial variable x j = concentration of j th species Y p = transformed factorial variable y = responses (composition, temperature, flame speed, critical times, etc.) in a chemically-reacting system α m = weighting factor for m th target in objective function β i = temperature exponent for i th forward reaction rate constant δ Bi = 1 if i th reaction contains species B; 0 otherwise η m = value of m th target in objective function ζ = vector of initial conditions θ = vector of parameters describing reacting system (rate constants, mass diffusivities, etc.) ν ij = signed stoichiometric coefficient of j th species in i th reaction ν′ ij , ν″ ij = forward and rever...

show abstract

Section: Methodsmentioning

confidence: 99%

Section: A Skeletal Model Generationmentioning

confidence: 99%

Section: A Skeletal Model Generationmentioning

confidence: 99%

“…In the previous study 16 the DRGEP method was used. For a kinetic model consisting of I reactions in J species, the rate of reaction of the i th reaction is:…”

Section: A Skeletal Model Generationmentioning

confidence: 99%