Session X40: Reacting Flows: General II

Understanding Critical Conditions for Wind-Driven Fire Spread Over a Fuel Gap

As wildfires increase in size and frequency, accurate predictions of fire behavior are needed to protect the lives of firefighters and mitigate losses in the wildland-urban interface (communities bordering natural lands). When responding to wildfires, firefighter safety zones are needed to provide a "fuel gap," ensuring safer operating conditions for firefighters in the event of unexpected fire behavior changes. Determining the appropriate size of these gaps to prevent injury under varying wind, fuel, and slope conditions is difficult because the relationship between heating and ignition is nonlinear and not well characterized. In this study, a series of spreading fire experiments were performed and analyzed at the USFS Missoula Fire Sciences Laboratory (wind speeds 0.5 - 1 m/s) and at the Insurance Institute of Business & Home Safety (wind speeds 6 - 9 m/s) to understand the critical conditions necessary for a spreading fire to "jump" across a fuel gap. These critical conditions are found to vary with wind speed and fuel properties, which together affect the size, geometry, and downstream heating of the fire approaching the gap. A scaling analysis based on the balance between momentum forces from the wind and buoyant forces from the flame described changes in flame geometry and downstream heating. This is correlated with the probability of ignition downstream. The coupling between fluid dynamics ahead of the flame and downstream ignition is further investigated and discussed.   

Characterization of a Novel Inclinable Wind Tunnel for the Fundamental Study of Wildfire Combustion

As the frequency and magnitude of wildfires grow worldwide, it is becoming increasingly important to understand the dynamics behind how these fires grow and spread. Wildfire spread is complex and affected by numerous variables. Two of these– cross-flow wind speed and ground slope– have a significant impact on the physical and chemical processes that occur during a burn. A novel inclinable wind tunnel facility (the WindCline) was constructed to study the coupling between the inertial cross-flow and buoyancy vectors in a controlled, combusting flow. The facility allows both the angle and ratio between the two vectors to be modified. The WindCline can operate at wind speeds up to 20 m/s with a freestream turbulence intensity less than 1% and is inclinable from -13 to +15 degrees. The 0.35 m wide by 0.85 m long test-section is highly modular to allow for combustion experiments with both solid fuel arrays and gaseous burners in flexible arrangements. The suite of measurement systems used to investigate both fluid dynamic and combustion processes in these flows includes high-speed particle image velocimetry, dual frequency comb laser spectroscopy, and particulate matter emissions sensors. A detailed characterization of the WindCline facility will be presented in addition to preliminary results for solid fuel combustion arrays of wooden pegs.   

Proper Orthogonal Decomposition of the Unperturbed Blue Whirl

The blue whirl (BW) is a recently discovered flame characterized by its blue color and minimal soot production. It was initially observed in laboratory-scale fire whirl (FW) experiments involving the combustion of liquid hydrocarbons on water (Xiao et al., PNAS, 2016); these BWs had diameters of 2-2.5 cm, heights of 6-8 cm, and were slightly lifted above water surface. Recent advancements in three-dimensional (3D) numerical simulations (Chung et al., Science Advances, 2020; Zhang et al., Computers and Fluids, 2018) have shed light on the flame structure. Now we conducted 3D direct numerical simulations of BWs, extending over 16 seconds of physical time and requiring an extremely refined computational grid. In addition, we used Proper Orthogonal Decomposition (POD) method to identify the primary flow structures and study flame and flow stability and dynamics.   

Reactive front propagation in turbulence

The relation between the propagation velocity of a thin flame and the turbulence intensity of the ambient flow is largely debated in the literature, giving rise to various parametrisations (e.g. [1-5]).

According to Damk ̈ohler’s analysis [1], based on the Huygens propagation model, a flame front propagates with a constant normal velocity even when wrinkled by turbulence. However, experiments using exothermic combustion reactions deviate from this model. The analysis of these experiments is complicated by the intermingling of turbulent and thermal effects, limiting potential improvement of the flame propagation parameterisation.

To test and improve Damk ̈ohler’s original theory, an autocatalytic chemical reaction is extensively studied experimentally and numerically. Analogous to a combustion reaction, it produces a thin reactive front separating the reactants from the products in an aqueous medium. Yet, this autocatalytic front does not show significant compressibility or thermal effects, and propagates at a much smaller velocity, hence facilitating metrology.  Accordingly, it allows to quantify the specific influence of turbulence on the speed and shape of the front in a refined framework.

In our experimental set-up, turbulence is generated by a system of oscillating grids establishing a homogeneous and isotropic turbulence in a closed water tank. Coupled measurements of PIV and LIF allow to follow, simultaneously, the velocity field and the propagation of the front.

In this presentation, the characteristics of the grid turbulence as well as the laminar propagation of the chemical reaction are first discussed and an estimated propagation law including the influence of chemical kinetics along with turbulence and its limitations are analyzed.

References

[1] Damköhler, G. Turbulence's influence on flame speed (1912).

[2] Schelkin K.I., Zhur. Techn. Fiz. 520 (no. 1110) (1943).

[3] Clavin, P. & Williams. F. A. Theory of premixed-flame propagation in turbulence (1979).

[4] Klimov, A.M. Interplay of hydrodynamics and chemistry in turbulent flames (1983).

[5]Gülder, ̈O. L. Turbulent flame propagation models for combustion regimes (1991).   

Direct Numerical Simulation of expansion, hydrodynamic mixing, and heat transfer to electrodes during low-temperature plasma discharges in atmospheric air

Low-temperature non-equilibrium plasmas have garnered significant interest from various communities. Nanosecond Pulsed Discharges (NSPD) in a pin-to-pin configuration are an efficient manner of generating such plasmas in ignition and flow control applications. However, the broad range of time and length scales involved in modelling NSPD configurations present formidable challenges to existing solvers. These challenges have led to the study of NSPDs in simplified conditions such as 0-D kinetics studies, the early discharge phase, or neglecting plasma processes altogether. In previous work, we presented a comprehensive solver that incorporates electrostatic processes, plasma relevant species, and the reactive Navier-Stokes equations in a fully-coupled manner. In this work, we use the solver to simulate a three-dimensional pin-to-pin NSPD in quiescent air at atmospheric conditions for several hundred microseconds. The asymmetric streamer propagation leads to the formation of a pressurized channel with two distinct temperature kernels near the electrodes. The pressurized channel relaxation results in complex shock dynamics for the following few microseconds, which is then followed by cool gas entrainment towards the middle of the gap. The metal electrode plays a prominent role both as a source of vorticity and heat transfer throughout the process. The observed phenomena are highly relevant to the induced hydrodynamics and mixing in the discharge channel, affecting subsequent pulses.   

Conditions for Satisfying the Second Law of Thermodynamics for Detailed and Reduced Chemical Mechanisms

In this study, we establish the necessary and sufficient conditions for the automatic satisfaction of the second law of thermodynamics in chemical reactions. We provide a proof of the sufficiency of these conditions for both real and ideal gases. Furthermore, we consider the impact of these conditions on common mechanism reduction methods. One such technique for mechanism reduction is to transform a system of N reversible reactions into 2N irreversible reactions, followed by the elimination of reactions based on specific criteria. We demonstrate that neglecting reactions in this way generally leads to violations of the second law. The magnitude of these violations vary, depending on the specific set of reactions and conditions. Consequently, we discuss the merits and drawbacks of common mechanism reduction techniques in light of the preceding analysis. Additionally, we explore the implication of these results on achieving thermodynamic consistency while employing global mechanisms.   

Quantifying the Relative Importance of Transport and Reaction Closures in a Canonical Premixed Turbulent Flow Setting

The study of reduced-order models for turbulent reacting flows often does not involve a disentangling of the underlying chemical and transport processes through a decoupling of the momentum and scalar equations. Previous work into the effects of turbulent transport on reaction dynamics and vice versa has shown that it is difficult to find analytic models for the prediction of mean scalar fields even in this simplified context. In this project, fully controlled and characterized homogeneous isotropic turbulence is used as a testbed to study closures involved in the evolution of a premixed reaction. Unresolved interactions inherent to incompressible turbulent flows are characterized across a broad parameter range and algebraic models are considered for closure terms. This analysis provides insight into the relative importance of modeling efforts for the scalar equation, which can be generalized to various heat source and reaction terms that support a propagating reaction front.   

Numerical study on the causality of solution multiplicity in ultra-lean H2-air premixed flames.

Safety issues in hydrogen-fueled devices can be related to flame ignition and propagation owing to leakages in non-ventilated gaps. In particular, large mass diffusivity of hydrogen enables the potential propagation of reacting kernels over very lean hydrogen-air mixtures in narrow channels h~O(1mm). Their formation and survival resides in the balance between the heat release due to combustion and the heat losses through the walls confining the lean mixture. Previous studies have found two stable isolated flame structures, an isolated circular flame and a double-headed one, that coexist under the same parametric space and propagate under different speeds [1]. This work presents a numerical study on the causality of solution multiplicity, which is thought to be determined by the symmetry-breaking details during ignition due to experimental evidence [2].

In particular, an ensemble of ignition cases including hot spots characterized by a temperature peak and linear ignition with a randomly sinusoidal morphology are investigated. We use a quasi-2D simplified formulation of the Navier-Stokes equations with a one-step Arrhenius reaction rate and off-plane heat losses, and an in-house Fourier-Fourier spectral discretization code to simulate the transient from ignition event.

A rich set of solutions evolving towards stable one and two-cell flames is generated for the same set of parameters. Flame structure tracking is proposed to study and characterize the formation of both configurations to unveil the underlying physics controlling the bistable behavior.   

On the occurrence of negative propagation speeds and negative divergence in turbulent premixed flames

Based on the analysis of data from recent Direct Numerical Simulation of turbulent premixed flames, at moderate values of the Reynolds number, we find that the propagation speed of reactive fronts in turbulent premixed flames takes instantaneous values that are negative in sign and large in absolute value. The ensuing probability density function (PDF) of the propagation speed is complex: the PDF has a positive mean, is bimodal with two local maxima – one at positive and one at negative values for sufficiently high values of the Reynolds number, is skew-positive, and has a large kurtosis indicating extreme events. Further, negative displacement speeds correlate strongly with negative values of the velocity divergence and large values of the most compressive eigenvalue of the rate of strain tensor. Despite strong similarities to isothermal turbulence, we also find an unmistakable dependence of the statistics of the rate of strain tensor on the Karlovitz number (at least for those values considered in our preliminary analysis), whereby as the Karlovitz number decreases, the probability that all strain eigenvalues are positive increases for small values of strain. The statistics of the propagation speed that we observe from DNS data are fundamentally inconsistent with existing theory that explains changes to propagation speed with the effect of stretch. Instead, the statistics of the propagation speed appear to be closely related to classical mechanisms of scalar mixing in isothermal turbulence with important Reynolds number effects pertaining to small-scale intermittency.