FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W37215635> ?p ?o ?g. }

• W37215635 abstract "In many conventional power plants, electricity is generated by burning natural gas in a gas turbine. Modern dry low-NOx combustion systems are efficient and have low emissions. Usually these combustion systems are operated in lean premixed combustion mode. In order to be able to operate a gas turbine as efficient as possible, to reduce emissions and to control flame stability, thorough understanding of the combustion process is required. The conversion rate of fuel in premixed combustion is determined by the burning speed. Under the highly turbulent conditions, as encountered in gas-turbine combustion, one can define a turbulent burning speed. The turbulent burning speed depends on the laminar burning speed, which strongly depends on the fuel. With developments towards a sustainable energy supply and the reduction of green- house gases, the use of alternative fuels is an option. In this context there is progress towards the application of clean coal technology and the use of biomass. Both meth- ods can result in the production of large amounts of hydrogen that can be used as a fuel. Replacing a part of the natural gas by hydrogen influences the chemical processes significantly, because hydrogen is relatively light and highly diffusive. This has an impact not only on the direct conversion rate of the fuel mixture but also on the sta- bility of the flame which can be responsible for an increased integral conversion rate. The phenomenon associated with the fact that all species have different diffusion rates, is referred to as preferential diffusion. In methane/air flames this effect is very small, but when hydrogen is present this effect is larger. Interactions of preferential diffusion and flame stretch, caused by e.g. turbulent flow, complicate things further. This makes it very worthwhile to study flame behavior of premixed stretched flames with hydrocarbon/hydrogen mixtures as a fuel. Numerical simulation is a powerful tool to gain insight in these complex combus- tion processes and can be used to predict what will happen under certain conditions. In the past only ad-hoc turbulent chemistry closures were used in which these pro- cesses could not be taken into account to determine the turbulent conversion rate. For fundamental investigations in which the interaction between small scale flow phenomena and the flame are important, all scales in the flow have to be resolved. Therefore, Direct Numerical Simulation (DNS) is the necessary tool to study turbu- lent flames. A drawback of DNS is that it is a computationally expensive method. In principle the chemical kinetics are taken into account by solving an equation for each species in the chemical system (detailed chemistry). This results in a very large and stiff system of equations and therefore the combination of DNS and detailed chem- istry is hardly feasible. Many studies use single-step chemistry in combination with DNS to limit the computational cost. The main drawback of using single-step chem- istry is that the complex effects due to multi-component flame behavior can not be taken into account. In this study, we investigate how the chemical reductionmethod Flamelet-Generated Manifolds (FGM) can be used to capture the chemical kinetics. In the FGM-approach the overall reaction progress is taken into account by solving an equation for one (or a few) reaction control variables. The FGM-approach reduces the number of equations to be solved and reduces the stiffness of the system of equations. Based on the reac- tion control variables, different quantities can be looked up in a flamelet database and can be used in the simulation. The combination of DNS and FGMis computationally feasible. Moreover, with a successful method the step towards statistical modeling with Large Eddy Simulations (LES) and Reynolds-Averaged Navier-Stokes (RANS) methods can be made for practical design calculations. The goal of the study presented in this thesis is to investigate the effects of flame stretch and preferential diffusion on the burning velocity of premixed flames. This will be done by using a DNS-FGM code. Detailed chemistry (which is feasible for one-dimensional cases) and single-step chemistry will be used to support results and to show differences between the different simulation approaches. By using detailed chemistry in one-dimensional stretched flames, it is found that not only the fuel but all chemical species in the system contribute to the flame (in)stability. The diffusivity of each species has an influence on the sensitivity of the mass burning rate to flame stretch. The sum of all these individual contributions results in the total sensitivity of the mass burning rate on the flame stretch. The re- sponse of the mass burning rate to flame stretch determines whether a stable or an unstable flame is found. In case of methane/air flames, different contributions can- cel and this results in a small net effect. The resulting sensitivity in ethane/air and propane/air flames is larger, the mass burning rate decreases more with the same flame stretch than a methane/air flame. Adding hydrogen to the fuel causes a de- creased sensitivity of the mass burning rate to flame stretch. As a first step in the application of FGMin a two-dimensional and three-dimensional computational domain, methane combustion is considered. In methane combustion there are no preferential diffusion effects. First we focus on differences between a single-step chemistry approach and a 1D-FGM (where only a single reaction control variable is used to look up the chemical source term). In the single-step chemistry approach the chemical source term distribution is functionally fixed and tuned to get a physical behavior. This is done in such a way that the correct flame thickness and flame speed are retained. With the application of FGM no tuning is required. It is shown that globally, for the applied coherent perturbations, no large differences exist between an FGM-approach and a single-step approach. In all flame-vortex in- teractions, the flame surface area and the mass consumption are slightly larger in the FGM-approach. In the turbulent flames small differences (positive and negative) are seen in the flame surface area and the mass consumption as well, however the differences are more or less random with time. Then, hydrogen is added to the fuel mixture and preferential diffusion effects cause local variation in stoichiometry and flame temperature. In order to describe the combustion reaction accurately, information on the local perturbed conditions is re- quired. The FGM-method is adjusted in such a way that an extra equation is solved that describes these local perturbations. It is clear that the effects can not be de- scribed by 1D methods, either single-step chemistry or a 1D-FGM. By application of a 2D-FGM, in which a constant flame stretch profile is taken into account, it is found that the results are much more accurate than a 1D-FGM simulation (or equiv- alently 1D single-step chemistry). This is concluded on the basis of a comparison in which several stretch fields along flamelet paths in the multi-dimensional solution are substituted in a 1D solver using detailed chemistry. It is concluded that the DNS- 2D-FGM numerical code is an efficient way to capture preferential diffusion effects to first order accuracy, taking (constant) stretch into account. This requires the use of two control variables. Adding hydrogen to the fuel results in a larger laminar mass consumption as well as in an increased turbulent consumption due to an increase in the flame surface area. The dimension of the FGM-database can be extended to gain accuracy, by taking stretch rate fluctuations through the flame front into account. In that case three control variables are needed. Progress is made in DNS with reduced chemistry for different hydrocarbons and methane/hydrogen mixtures under perturbed flow conditions, yielding insight in the flame behavior. This is essential for model closure in simulations of flame behav- ior in practical applications. The step towards application in LES/RANS for more practical large scale design calculations needs further research, especially in the area of the behavior of (subgrid) fluctuations in the presence of instabilities." @default.
• W37215635 created "2016-06-24" @default.
• W37215635 creator A5061157489 @default.
• W37215635 date "2009-01-01" @default.
• W37215635 modified "2023-09-23" @default.
• W37215635 title "Modeling and analysis of flame stretch and preferential diffusion in premixed flames" @default.
• W37215635 doi "https://doi.org/10.6100/ir651971" @default.
• W37215635 hasPublicationYear "2009" @default.
• W37215635 type Work @default.
• W37215635 sameAs 37215635 @default.
• W37215635 citedByCount "1" @default.
• W37215635 countsByYear W372156352013 @default.
• W37215635 crossrefType "journal-article" @default.
• W37215635 hasAuthorship W37215635A5061157489 @default.
• W37215635 hasConcept C105923489 @default.
• W37215635 hasConcept C121332964 @default.
• W37215635 hasConcept C129474609 @default.
• W37215635 hasConcept C178790620 @default.
• W37215635 hasConcept C185592680 @default.
• W37215635 hasConcept C192562407 @default.
• W37215635 hasConcept C57879066 @default.
• W37215635 hasConcept C6506403 @default.
• W37215635 hasConcept C69357855 @default.
• W37215635 hasConcept C83104080 @default.
• W37215635 hasConcept C97355855 @default.
• W37215635 hasConceptScore W37215635C105923489 @default.
• W37215635 hasConceptScore W37215635C121332964 @default.
• W37215635 hasConceptScore W37215635C129474609 @default.
• W37215635 hasConceptScore W37215635C178790620 @default.
• W37215635 hasConceptScore W37215635C185592680 @default.
• W37215635 hasConceptScore W37215635C192562407 @default.
• W37215635 hasConceptScore W37215635C57879066 @default.
• W37215635 hasConceptScore W37215635C6506403 @default.
• W37215635 hasConceptScore W37215635C69357855 @default.
• W37215635 hasConceptScore W37215635C83104080 @default.
• W37215635 hasConceptScore W37215635C97355855 @default.
• W37215635 hasLocation W372156351 @default.
• W37215635 hasOpenAccess W37215635 @default.
• W37215635 hasPrimaryLocation W372156351 @default.
• W37215635 hasRelatedWork W1535450206 @default.
• W37215635 hasRelatedWork W173777646 @default.
• W37215635 hasRelatedWork W1898329713 @default.
• W37215635 hasRelatedWork W2005323845 @default.
• W37215635 hasRelatedWork W2102271666 @default.
• W37215635 hasRelatedWork W2104787903 @default.
• W37215635 hasRelatedWork W2154280661 @default.
• W37215635 hasRelatedWork W2238103014 @default.
• W37215635 hasRelatedWork W2238762106 @default.
• W37215635 hasRelatedWork W2520771902 @default.
• W37215635 hasRelatedWork W2611866365 @default.
• W37215635 hasRelatedWork W2728531517 @default.
• W37215635 hasRelatedWork W2774313707 @default.
• W37215635 hasRelatedWork W2803239943 @default.
• W37215635 hasRelatedWork W2899176439 @default.
• W37215635 hasRelatedWork W2916043084 @default.
• W37215635 hasRelatedWork W2996564198 @default.
• W37215635 hasRelatedWork W3123166357 @default.
• W37215635 hasRelatedWork W658835224 @default.
• W37215635 hasRelatedWork W88156014 @default.
• W37215635 isParatext "false" @default.
• W37215635 isRetracted "false" @default.
• W37215635 magId "37215635" @default.
• W37215635 workType "article" @default.