Session E09: Reacting Flows: Modeling and Simulation I

Adjoint-based sensitivity of turbulent reacting jets

In this talk, we present a discrete adjoint-based method for measuring local sensitivity in direct numerical simulations of turbulent reacting jets. We introduce a novel adaptive dissipation scheme compatible with the discrete adjoint formulation that preserves scalar boundedness while retaining high-order accuracy. A flamelet-progress variable approach is employed to handle chemical reactions using tabulated chemistry. The utility of using tabulated chemistry is that discrete adjoint sensitivity can be computed efficiently for arbitrary chemical mechanisms. Putting this together, direct numerical simulations of a three-dimensional reacting round jet are performed at moderate Reynolds numbers. Sensitivity obtained from the adjoint solution is used to study flame and chemistry responses to turbulent flow actuation.   

Not Participating

A novel molecular simulation technique, called the Direct Simulation Monte Carlo (DSMC) method, is used to simulate one-dimensional hydrogen-air flames. In this method, the dynamics of the real molecules are emulated with the help of the simulation particles, and their averages provide the macroscopic flow properties. The Quantum-Kinetic (QK) model, which uses total collision energy, molecular dissociation energy, quantized vibrational energy levels and the principle of molecular reversibility, is employed to account for the reactions. A detailed 21-step reaction mechanism is utilised for describing combustion. Systematic variations in the volume fraction of hydrogen are performed in order to evaluate the flammability limit of the hydrogen-air mixture as well as the corresponding flame speed and flame temperature. The results are compared with the experimental data from the literature. The hydrogen volume fractions corresponding to the lean and rich flammability limits are found to match closely with those from previous experimental data. Also, the temperature and flame speeds for the range of cases show good agreement with the measurements. However, some differences are still present which can be mitigated as more accurate molecular-level and/or ab-initio data become available. A distinct advantage of DSMC is that, because it uses molecular level simulations, it avoids using continuum Arrhenius reaction rates and simplified gradient laws for the diffusivities of mass, momentum and heat. This technique is useful in cases when the continuum assumptions or empirical formulations do not hold true.    

The effect of mixing of miscible liquids with disparate viscosity in co-axial flow on chemical reactions

Achieving optimal process results in mixing limited – fast reactions is a long-standing challenge in chemical industry. For complex reactions with multiple desirable and undesirable products, solutions to optimizing the product yield could lead to inefficiencies, such as reducing productivity, increasing solvent/dilution or reduced catalyst activity and so on. Providing improvements to mixing in such systems becomes further complicated if there are large viscosity differences between the reacting fluids.

This talk will discuss the effect of disparate viscosity fluid mixing on fast chemical reactions in co-axial jets. High fidelity experiments and numerical simulations show that the viscosity ratio between the two reacting streams lead to significant differences in mixing and therefore in resultant chemical species. A novel in-line spectrometry technique has been developed and validated to measure the resultant chemical species in the product stream. Numerical modeling of chemical reactions is based on the Li &Toor [Li & Toor, AIChE J, 32(8)1986] approach. The comparison of the numerical modeling to the experimental data highlight the importance of mixing on pH and therefore on the chemical kinetics.   

Conditional momentum equation for modeling heat release effects on turbulence in turbulent premixed combustion

Closure models for the conditionally-averaged momentum transport equation with respect to a progress variable have recently been developed to capture the influence of combustion heat release on turbulence in turbulent premixed combustion. In this work, a posteriori validation is performed in a spatially-evolving turbulent premixed hydrogen/air planar jet flame where heat release effects on turbulence are significant. For evaluating the conditional analog of the turbulent viscosity, a model conditional turbulent kinetic energy transport equation is solved, and an algebraic model of the conditional turbulent dissipation is utilized. Compared to conventional turbulence models, the conditional momentum equation is shown to correctly predict counter-gradient transport in low Karlovitz number turbulent premixed combustion without any a priori specification that such a phenomena should occur.   

Scalar transport closure for a nonlinear reaction problem

Building reduced-order models for turbulent reacting flows is theoretically and computationally challenging as the underlying chemical and transport processes are individually complex and a thorough understanding of the coupled effects of these phenomena remains elusive. Deeper insight into the effects of turbulent transport on reaction dynamics and vice versa is essential for the future design of engineering systems, including those for efficient energy conversion. In this work, a reacting scalar problem that mimics flamelet behavior for premixed combustion is studied in a canonical mixing configuration. The closure terms and closure operators in the context of the Reynolds-Averaged Navier-Stokes equations are developed. We present an assessment of the local and non-local forms of the closure operators capable of capturing unresolved interactions between scalar reactions, building on similar work for non-premixed combustion.    

One-dimensional simulations of multicomponent counterflow spray flames

Many modern combustion systems such as gas turbines and rocket engines rely on spray combustion, yet this process is not fully understood due to its complex physical processes, involving convective, diffusive, chemical, and multiphase effects. In particular, the effects of preferential evaporation are significant for transportation fuels, which consist of a large number of individual compounds. This research presents numerical investigations of a counterflow spray flame using a one-dimensional solver. An Eulerian-Eulerian approach is employed in which the spray is modeled as a monodisperse, continuous diffuse phase. This technique is used to simulate flames with multicomponent fuels using both preferential and mixture-averaged evaporation models. The effects of preferential evaporation on flame structure are presented via construction of regime diagrams with respect to droplet diameter and strain rate.   

Not Participating

Lean highly-diffusive mixtures of hydrogen-air configure a reactive media that enables the potential propagation of reacting kernels in gaps exposed to fuel leakage. In this work, a study on nearly-two-dimensional reacting fronts under canonical configurations is proposed.

A flame sheet model is used to elucidate the mathematical existence of steady propagating solutions of temperature and reacting species fields related to semicircular flames. To this end, two physical mechanisms are proposed, convection and local heat losses, which enable the otherwise unattainable planar canonical solutions in a quiescent environment. Theoretical base steady flows in the flame-sheet limit enable subsequent studies on the stability and evolution of these type of structures.

Numerical validation is achieved via an in-house finite-element code for reactive flows in the low-Mach approximation. A quasi-2D description of the conservation equations with a one-step Arrhenius chemical model and off-the-plane conductive heat losses is used to simulate lean hydrogen-air premixed flames that slowly propagate against weak convective streams of reactants. In particular, two-headed and semicircular flames are recovered as heat losses are increased.   

Models for Flame-Vortex-Acoustic Interactions in a Backward-Facing Step Combustor - A Comparitive Study

A novel model for flame-vortex-acoustic interactions in a lean premixed step combustor is described. The distinctive role of vortex shedding motivates the use of a Lagrangian vortex method for the flow. Two flame models are compared, the G-Equation (GE) and Thermo-Diffusive model (TD). At stable operation, weak wrinkles propagate along the flame surface in both models and their instantaneous flame positions match. For larger equivalence ratios, a higher heat release perturbation q' causes a transition to limit-cycle behaviour in the TD where flame-vortex interactions over one cycle are predicted. When the velocity reaches a minimum, vortices in the recirculation zone merge with vortices at step upstream increasing the circulation Γ resulting in a large vortex that extends to the top wall. The flame wraps around the large vortex trapping reactants in its fold. The trapped reactants burn vigorously, marking the instant of highest q' in phase with acoustics-forced Γ. This contributes to the closed-loop feedback mechanism. In the GE, high u_θ/s_L due to strong Γ tend to cause large wrinkles and cusps that do not smooth out due to absence of kinematic restoration effects, otherwise captured by the TD. The TD dynamics and phase relationships are in excellent agreement with experiments.   

Not Participating

Reactive flows are the flows involving fluids undergoing a chemical reaction when in contact. We explore the reactive flows in porous medium in connection with the enhanced oil recovery, CO₂ sequestration, fluid mixing, to name a few. We consider miscible fluids undergoing radial displacement and a second order chemical reaction to generate a product. We model the flow using Darcy’s law coupled to a system of convection-diffusion-reaction equations. A hybrid numerical scheme based on compact finite difference and pseudo-spectral method is used to understand various scalings of the amount of product generated and the spreading of the product with time. The product having different viscosity than the reactants results a hydrodynamic instability called viscous fingering (VF). The deformation of the fluid-fluid interface into finger-like patterns when a less viscous fluid displaces a more viscous fluid is termed as VF. A novel linear stability analysis is performed to understand the onset of the instability wherein similarity and differences with rectilinear displacement are discussed.     

Simulating wildfires from single branches to local weather systems

Amplified by climate change, wildfires are causing devastating damages and loss of life, and are indeed becoming an existential threat in some regions of the U.S. and worldwide. Accurate simulations of wildfires can be used in wildfire prevention, in training, and for operational planning during an ongoing wildfire. Current fast methods use concepts of reaction–diffusion systems, wave proagation, or percolation, whereas methods that model individual trees typically take significant computational time on supercomputers. We present a new numerical multiscale framework that models tree structures down to branches, includes fluid dynamical modeling of local weather systems over realistic terrain, and runs at interactive rates. Our method allows tree representation with a previously unavailable fidelity, while simultaneously capturing atmospheric fluid dynamics including flammagenitus cloud formation and precipitation caused by the fire. The method captures the impact of tree topology, terrain characteristics, and ecosystem composition on fire spread. The interactive computational speed allows for an intuitive interaction with the simulation, where the impact of fire retardants, firebreaks, or rain can be experienced and studied.