Search

We report a cascaded group-additivity (CGA) ONIOM method for high-level energy calculations of large aliphatic hydrocarbon molecules by combining the group additivity and two-layer ONIOM methods. This hybrid method is implemented by partitioning the target molecule into individual groups, which are cascaded via the overlapping between them. The energy of the entire molecule is first calculated at a low level of theory such as M06-2x/cc-pVTZ. Then all the groups and their overlappings are treated at the levels of CCSD(T)/CBS and M06-2x/cc-pVTZ to obtain their energy difference to be used as the energy correction. We selected small-to-middle size aliphatic hydrocarbons including 79 C4 C8 molecules as the validation set to demonstrate the feasibility of the CGA-ONIOM method, followed by the calculations of 12 representative C10, C12 and C16 aliphatic hydrocarbons (including normal-, branched-, cyclo- and unsaturated categories). Our calculations agree well with the reference values available in the literature with the modest deviation around 1.0 kcal mol−1. Compared with the conventional CCSD(T)/CBS calculation of the whole molecule, the computational cost can be dramatically reduced by a factor of ∼102 for molecules with 10 carbons and ∼104 for molecules with 16 carbons. Considering its outstanding computational efficiency and accuracy, our proposed CGA-ONIOM method is promising for combustion chemistry studies of large fuel molecules at a high level of theory.

The formulation of reactive flow problems can be both quite challenging and important to the efficiency and robustness of solution algorithms. In this article, we focus on the choice of the thermochemical state vector as it relates to recently-developed computational techniques for complex combustion chemistry problems. We identify over-specification of the state vector as a source of both ambiguity and error in the partial derivatives used in forming analytical forms of the chemical source Jacobian matrix. We review and compare several approaches taken to increase sparsity of the Jacobian matrix, as it relates to the use of Newton–Krylov methods for implicit time integration, and identify proper techniques for achieving sparsity that do not rely on ad-hoc choice of state variables with inconsistent Jacobians. Chemical explosive mode analysis and linearly-implicit methods, such as Rosenbrock methods, are identified as areas where Jacobian accuracy may be critical. The distinction between how Jacobian exactness impacts Rosenbrock and Newton–Krylov methods is demonstrated with a simple example. We demonstrate the errors in conservation obtained from over-specification of the state vector with auto-ignition calculations for hydrogen, ethylene, and n-heptane chemistry with a high-order implicit Runge–Kutta method.

In this study, a series of ignition delay time (IDT) experiments of 1-butene were performed in a high-pressure shock tube (HPST) and in a rapid compression machine (RCM) under conditions of relevance to practical combustors. This is the first 1-butene IDT data taken at engine relevant conditions, and the combination of HPST and RCM results greatly expands the range of data available for the oxidation of 1-butene to higher pressures (10-50 atm), lower temperatures (670-1350 K) and to a wide range of equivalence ratios (0.5-2.0).A comprehensive chemical kinetic mechanism to describe the combustion of 1-butene has simultaneously been applied. It has been validated using the IDT data measured here in addition to a large variety of literature data: IDTs, speciation data from jet-stirred reactor (JSR), premixed flame, and flow reactor, and laminar flame speed data. Important reactions have been identified via flux and sensitivity analyses including: (a) H-atom abstraction from 1-butene by hydroxyl radicals and molecular oxygen from different carbon sites; (b) addition reactions, including hydrogen atom and hydroxyl radical addition to 1-butene; (c) allylic radical chemistry, including the addition reactions with methyl radical, hydroperoxy radical and self-recombination; (d) vinylic radical chemistry, including the addition reaction with molecular oxygen; (e) alcohol radical chemistry, including the Waddington type propagating reaction pathways and alkyl radical low-temperature branching chemical pathways. (C) 2017 The Combustion Institute. Published by Elsevier Inc. All rights reserved. The authors thank the entire group members at Combustion Chemistry Centre for helpful discussions. This work at NUI Galway was supported by Saudi Aramco under the FUELCOM program. peer-reviewed 2019-04-13

International audience; Fuels of the furan family, i.e. furan itself, 2-methylfuran (MF), and 2,5-dimethylfuran (DMF) are being proposed as alternatives to hydrocarbon fuels and are potentially accessible from cellulosic biomass. While some experiments and modeling results are becoming available for each of these fuels, a comprehensive experimental and modeling analysis of the three fuels under the same conditions, simulated using the same chemical reaction model, has - to the best of our knowledge - not been attempted before. The present series of three papers, detailing the results obtained in flat flames for each of the three fuels separately, reports experimental data and explores their combustion chemistry using kinetic modeling. The first part of this series focuses on the chemistry of low-pressure furan flames. Two laminar premixed low-pressure (20 and 40 mbar) flat argon-diluted (50%) flames of furan were studied at two equivalence ratios (phi = 1.0 and 1.7) using an analytical combination of high-resolution electron-ionization molecular-beam mass spectrometry (EI-MBMS) in Bielefeld and gas chromatography (GC) in Nancy. The time-of-flight MBMS with its high mass resolution enables the detection of both stable and reactive species, while the gas chromatograph permits the separation of isomers. Mole fractions of reactants, products, and stable and radical intermediates were measured as a function of the distance to the burner. A single kinetic model was used to predict the flame structure of the three fuels: furan (in this paper), 2-methylfuran (in Part II), and 2,5-dimethylfuran (in Part III). A refined sub-mechanism for furan combustion, based on the work of Tian et al. [Combust. Flame 158 (2011)756-773] was developed which was then compared to the present experimental results. Overall, the agreement is encouraging. The main reaction pathways involved in furan combustion were delineated computing the rates of formation and consumption of all species. It is seen that the predominant furan consumption pathway is initiated by H-addition on the carbon atom neighboring the O-atom with acetylene as one of the dominant products.

There is much interest in the combustion mechanism of organophosphorus compounds (OPCs) due to their role as potential halon replacements in fire suppression. A continuing investigation of the inhibition activity of organophosphorus compounds under a range of equivalence ratios was performed experimentally and computationally, as measured by the burning velocity. Updates to a previous mechanism were made by the addition and modification of reactions in the mechanism for a more complete description of the inhibition reactions. Reaction pathways for HOPO 2 + H and HOPO + H are analyzed using the BAC-G2 approach. A new reaction pathway for HOPO 2 + H = PO 2 + H 2 O has been identified which results in a higher rate constant than that reported in the literature. In this work, the laminar flame speed is measured experimentally and calculated numerically for a premixed propane/air flame at 1 atm, under a range of equivalence ratios, undoped and doped with dimethyl methylphosphonate (DMMP). A detailed investigation of the catalytic cycles involved in the recombination of key flame radicals is made for two equivalence ratios, fuel lean and fuel rich. From this, the importance of different catalytic cycles involved in the lean versus rich case is discussed. The chemical kinetic model indicates that the HOPO 2 ⇔ PO 2 inhibition cycle is more important in the lean flame than the rich. The OPCs are similarly effective across the range, demonstrating the robustness of OPCs as flame suppressants. In addition, it is shown that the phosphorus compounds are most active in the high-temperature region of the flame. This may, in part, explain their high level of inhibition effectiveness.

Combustion mechanisms of higher hydrocarbons are governed by a hierarchical structure. This structure is reflected in reduced mechanisms based on intrinsic low dimensional manifolds (ILDM). Based on the mathematical analysis we present in this work a hierarchical concept where the ILDMs of higher hydrocarbons are generated on the basis of existing ILDMs of simpler reaction mechanisms. It is based on a numerical method which allows to estimate, how well an ILDM can be represented by an ILDM of a simpler subsystem. If the conditions for a good representation of the ILDM by the simpler subsystem are met, then the ILDM of the simpler system can be used directly. If not, then the ILDM of the simpler subsystem is used as a starting estimate for the generation of the ILDM which reduces the computational effort considerably. The example of a syngas ILDM which is used as a generic ILDM for n-heptane combustion validates the approach.

International audience; This work is the third part of a study focusing on the combustion chemistry and flame structure of furan and selected alkylated derivatives, i.e. furan in Part I, 2-methylfuran (MF) in Part II, and 2,5-dimethylfuran (DMF) in the present work. Two premixed low-pressure (20 and 40 mbar) flat argon-diluted (50%) flames of DMF were studied with electron-ionization molecular-beam mass spectrometry (EI-MBMS) and gas chromatography (GC) under two equivalence ratios (phi = 1.0 and 1.7). Mole fractions of reactants, products, and stable and radical intermediates were measured as a function of the distance to the burner. Kinetic modeling was performed using a reaction mechanism that was further developed in the present series, including Part I and Part II. A reasonable agreement between the present experimental results and the simulation is observed. The main reaction pathways of DMF consumption were derived from a reaction flow analysis. Also, a comparison of the key features for the three flames is presented, as well as a comparison between these flames of furanic compounds and those of other fuels. An a priori surprising ability of DMF to form soot precursors (e.g. 1,3-cyclopentadiene or benzene) compared to less substituted furans and to other fuels has been experimentally observed and is well explained in the model.

Laser-induced fluorescence (LIF) provides a selective, sensitive, and nonintrusive means of measuring reactive intermediates in combustion chemistry. This article examines ways in which LIF can be used in a semiquantitative manner to gain insight into certain features of flame chemistry mechanisms, through concentration estimates accurate to within perhaps a factor of three. Experiments detecting NH and NS are described, and extensions of semiquantitative LIF to multiple-species detection including two-dimensional imaging are discussed.