An integrated data-driven workflow for kinetic model development and optimisation: theory and application to OMEs combustion

Flame structure and a compact reaction mechanism for combustion of dimethyl ether at atmospheric pressure

In direct numerical simulations of reactive flow and other multi scale physical problems, a broad time scale distribution from seconds to a fraction of nano-seconds makes explicit schemes in-efficient. On the other hand, although implicit methods may improve computational efficiency, they lose the time histories and statistical values of the fast modes, which may be required for modeling of large scale slow modes in a turbulent reactive flow and multi physical process. In the present study, a dynamic multi time scale (MTS) method and a dynamic hybrid multi time scale (HMTS) model are developed to achieve efficient and time accurate integration of stiff ODEs with detailed kinetic mechanisms. The methods are applied to ignition of hydrogen, methane and n-decane/air mixtures and compared, respectively, with standard Euler and implicit ODE solvers by using detailed chemical kinetic mechanisms. The results showed that both methods can accurately reproduce the species time histories and ignition delay times. In addition, compared to the explicit Euler method, MTS is not only computationally efficient but also robust at larger time steps. Compared to the implicit ODE solver, MTS is about one-order more computationally efficient. In addition, unlike the implicit ODE solver, whose computation time is proportional to the square of the species number, the computation time required for MTS is only proportional linearly to the species number. As such, MTS has advantages particularly for large equation systems such as large chemical kinetic mechanisms. To further accommodate the specification of a limiting time scale of the equation system and to improve the computation efficiency and robustness at large time scales, HMTS is developed by integrating MTS with a fully implicit algorithm. Therefore, the present HMTS is a generalized scheme which includes the Euler scheme, MTS, and implicit scheme, and compatible to both incompressible and compressible flow solvers. The results showed HMTS is rigorous and efficient. This scheme can be used for direct numerical simulations and large eddy simulation with detailed chemical mechanisms to improve the computation efficiency, accuracy, and robustness.

: The VULCAN CFD code integrated with a reduced chemical kinetic mechanism was applied to simulate cavity-stabilized ethylene-air flames and to predict flame stability limits in supersonic flows based on an experimental study. A 15-step reduced kinetic mechanism for ethylene was systematically developed through skeletal reduction with a directed relation graph and time scale reduction based on quasi-steady state assumptions. The accuracy of the reduced kinetic mechanism and its implementation in the VULCAN code were demonstrated in an auto-ignition problem with a range of parameters. 3D simulations were then carried out for cavity-stabilized flames at different fuel flow rates and turbulent Schmidt numbers. For comparison with the performance of the present reduced mechanism, a 3- and a 10-step global kinetic model were applied to simulate the same cavity combustor, and the results show that the 15-step reduced model predicts experimental results much better than the 3- and 10-step models. The importance of including accurate chemical kinetics in CFD simulations is therefore demonstrated.

Chemistry acceleration with tabulated dynamic adaptive chemistry in a realistic engine with a primary reference fuel

The positive effect of plasma discharges on ignition and flame stability motivates the development of detailed kinetic mechanisms for high-fidelity simulations of plasma-assisted combustion. Because of their hierarchical nature, combustion processes require a large number of chemical species and pathways to describe hydrocarbon oxidation. In order to simulate kinetic enhancement by non-thermal electrons, additional species and processes are included, which model the ionization and excitation of neutral molecules. From a practical perspective, integrating large kinetics mechanisms is computationally burdensome due to the temporal stiffness of the nonlinear combustion dynamics and the memory requirements associated with the high number of species. In order to alleviate computational costs, a dimensionality reduction approach is proposed based on principal component analysis. The methodology is demonstrated on a detailed kinetics mechanism for plasma-assisted combustion excited by a nanosecond pulse discharge. Data are collected from a zero-dimensional two-temperature reactor model, whereby a nanosecond pulse generates a population of excited-state molecules and radicals in argon and air mixtures with hydrocarbon fuels. The data from the detailed mechanism are used to describe the evolution of the plasma and mixture based on principal components. Several skeletal mechanisms consisting of a much smaller number of species are assembled and their accuracy is compared against the detailed one. The performance of selected skeletal mechanisms is found satisfactory for the simulation of plasma-assisted ignition in unsteady, three-dimensional reactive flows.

An extensive study on skeletal mechanism reduction for the oxidation of C0–C4 fuels

The finite volume method (FVM) is widely recognized as a computationally efficient and accurate mesh-based technique. However, it has notable limitations, particularly in mesh generation and handling complex boundary interfaces or conditions. In contrast, the smoothed particle hydrodynamics (SPH) method, a popular meshless alternative, inherently circumvents the challenges of mesh generation and yields smoother numerical outcomes. Nevertheless, this approach comes at the cost of reduced computational efficiency. Consequently, researchers have strategically combined the strengths of both methods to investigate complex flow phenomena, producing precise and computationally efficient outcomes. However, algorithms involving the weak coupling of these two methods tend to be intricate and face challenges regarding versatility, implementation, and mutual adaptation to hardware and coding structures. Thus, achieving a robust and strong coupling of FVM and SPH within a unified framework is essential. A mesh-based FVM has recently been integrated into the SPH-based library SPHinXsys. However, due to the differing boundary algorithms between these methods, the crucial step for establishing a strong coupling of both methods within a unified SPH framework is to incorporate the FVM boundary algorithm into the Eulerian SPH method. In this paper, we propose a straightforward algorithm within the Eulerian SPH method, which is algorithmically equivalent to that in FVM and based on the principle of zero-order consistency. Moreover, several numerical examples, including compressible and incompressible flows with various boundary conditions in the Eulerian SPH method, demonstrate the stability and accuracy of the proposed algorithm.

Multi-timescale and correlated dynamic adaptive chemistry modeling of ignition and flame propagation using a real jet fuel surrogate model

A dynamic multi-timescale method for combustion modeling with detailed and reduced chemical kinetic mechanisms

Skeletal reaction models are derived for a four-component gasoline surrogate model via a novel fully automated instantaneous local sensitivity analysis technique. This technique provides a direct approach for model reduction of large systems; in contrast to previous/ current methods constructed by ‘hands-on manipulation.’ The sensitivities of the species mass fractions and the temperature with respect to the reaction rates are estimated by a reduced-order modelling (ROM) methodology. Termed ‘implicit time-dependent basis CUR (implicit TDB-CUR),’ this methodology is based on the CUR matrix decomposition and incorporates implicit time integration to evolve the bases. The estimated sensitivities are subsequently analysed to develop skeletal reaction models with a fully automated procedure. The 1389-species gasoline surrogate model developed at Lawrence Livermore National Laboratory (LLNL) is selected as the detailed kinetics model. The skeletal reduction procedure is applied to this model in a zero-dimensional constant-pressure reactor over a wide range of initial conditions. The performances of the resulting models are evaluated by comparison against the results produced by the LLNL detailed model and also predictions by other models. Two new skeletal models consisting of 679 and 494 species are developed. The first is an alternative to an existing model with the same number of species. The predictions with this model reproduce the flame results obtained with the detailed model with less than 1% errors. The errors via the second model are less than 10%.

A path flux analysis method for the reduction of detailed chemical kinetic mechanisms

Multiscale modeling and simulation methods for electromagnetic and multiphysics problems

Multi-Object tracking in real world environments is a tough problem, especially for cell morphogenesis with division. Most cell tracking methods are hard to achieve reliable mitosis detection, efficient inter-frame matching, and accurate state estimation simultaneously within a unified tracking framework. In this paper, we propose a novel unified framework that leverages a spatio-temporal ant colony evolutionary algorithm to track cells amidst mitosis under measurement uncertainty. Each Bernoulli ant colony representing a migrating cell is able to capture the occurrence of mitosis through the proposed Isolation Random Forest (IRF)-assisted temporal mitosis detection algorithm with the assumption that mitotic cells exhibit unique spatio-temporal features different from non-mitotic ones. Guided by prediction of a division event, multiple ant colonies evolve between consecutive frames according to an augmented assignment matrix solved by the extended Hungarian method. To handle dense cell populations, an efficient group partition between cells and measurements is exploited, which enables multiple assignment tasks to be executed in parallel with a reduction in matrix dimension. After inter-frame traversing, the ant colony transitions to a foraging stage in which it begins approximating the Bernoulli parameter to estimate cell state by iteratively updating its pheromone field. Experiments on multi-cell tracking in the presence of cell mitosis and morphological changes are conducted, and the results demonstrate that the proposed method outperforms state-of-the-art approaches, striking a balance between accuracy and computational efficiency.

Recently, coinciding with and perhaps driving the increased popularity of prediction markets, several novel pari-mutuel mechanisms have been developed such as the logarithmic market scoring rule (LMSR), the cost-function formulation of market makers, and the sequential convex parimutuel mechanism (SCPM). In this work, we present a unified convex optimization framework which connects these seemingly unrelated models for centrally organizing contingent claims markets. The existing mechanisms can be expressed in our unified framework using classic utility functions. We also show that this framework is equivalent to a convex risk minimization model for the market maker. This facilitates a better understanding of the risk attitudes adopted by various mechanisms. The utility framework also leads to easy implementation since we can now find the useful cost function of a market maker in polynomial time through the solution of a simple convex optimization problem.In addition to unifying and explaining the existing mechanisms, we use the generalized framework to derive necessary and sufficient conditions for many desirable properties of a prediction market mechanism such as proper scoring, truthful bidding (in a myopic sense), efficient computation, controllable risk-measure, and guarantees on the worst-case loss. As a result, we develop the first proper, truthful, risk controlled, loss-bounded (in number of states) mechanism; none of the previously proposed mechanisms possessed all these properties simultaneously. Thus, our work could provide an effective tool for designing new market mechanisms.

In this dissertation a generalized kinetic mechanism for photochemical smog is formulated and validated. There are two basic objectives for pursuing this work. First, kinetic mechanisms are a critical component of airshed simulation models whose uses include the evaluation of alternative control strategies for photochemical smog. Second, kinetic mechanisms provide a means of understanding the chemistry of smog formation. In addition to developing the kinetic mechanism, extensive consideration is given here to further experimental studies of the kinetics of elementary reactions and to procedural aspects of smog chamber experiments which will aid in reducing uncertainty in future mathematical simulation studies. The most important feature of the kinetic mechanism which 1s presented in Chapter I lies in its general nature; that is, the mechanism has been written so as to be applicable to a large number of hydrocarbons -- and, ultimately, the entire atmospheric hydrocarbon mix -- rather than just a specific hydrocarbon such as propylene. The rationale for the lumping procedure is described in detail. By design the resultant mechanism takes advantage of the general features typical of smog formation to maintain at a minimum the number of reactions and species included while at the same time retaining a high degree of detail, especially as concerns the chemistry of the inorganic species. In essence a careful balance between compactness of form and accuracy of prediction is sought. The mechanism is then validated using n-butane-NOx, propylene-NOx, and n-butane-propylene- NOx smog chamber data at 13 different sets of initial reactant concentrations and a wide variety of hydrocarbon to NOx ratios. Several sources of potential uncertainty in the predictions of the mechanism are discussed. Of these the two most serious -- and the two most amenable to correction -- are gaps in our knowledge of rate constants and mechanisms of key elementary reactions and effects related to smog chamber systems which either alter or incorrectly monitor the course of smog formation in controlled experimenta1 studies. These sources of uncertainty along with recommendations for future studies to minimize them are the topics of Chapters II and III.