Forman A. Williams

Numerical simulations are conducted to study fundamental aspects of combustion stabilization in hydrogen-fueled gas turbines. The study focuses on laminar lifted flames at moderate Reynolds numbers in axisymmetric configurations, where a swirling hydrogen jet diluted with nitrogen is injected into stagnant, preheated, pre-compressed air. The conservation equations are formulated in the low-Mach-number approximation, employing a mixture-averaged model for molecular transport. Fuel oxidation is modeled using both detailed chemical kinetics and a previously derived explicit one-step reduced mechanism, which assumes steady-state behavior for chemical intermediates—a valid approximation under the high-pressure conditions typical of gas-turbine combustion chambers, and the accuracy of that approximation is ascertained. The investigation explores the interplay between vortex breakdown and flame dynamics, including liftoff and blowoff, as functions of the swirl and Damköhler numbers. The results elucidate the required flow criteria for lifted-flame stabilization and demonstrate the predictive capability and computational cost reduction of the one-step chemistry in connection with hydrogen combustion at high pressures. A regime diagram in a plane of swirl number and Damköhler number is derived, and conditions for the occurrence of steadily pulsating flames are established, along with indications of amplitudes and frequencies of those oscillations. While clearly not directly applicable to practical turbulent-flow conditions, the results can be useful in future analyses and design concepts for combustion chambers of hydrogen-fueled gas turbines. Novelty and significance statement This work presents, for the first time, results of computations of nitrogen-diluted hydrogen flame behavior for swirling fuel jets issuing into air that has been heated to temperatures expected at the entrance to gas-turbine combustion chambers. It is novel in that it compares predictions made using both detailed combustion chemistry and one-step systematically derived reduced chemistry. A significant finding is that the results obtained with the reduced chemistry are in general agreement with those of the detailed chemistry, thereby affording substantial reductions in computational cost. Another novel and significant result is the determination of injection and swirl gas-turbine conditions required for stable lifted flames to occur, rather than attached flames or blowoff. The existence and characteristics of pulsating oscillations also are established for the first time. These results will be useful in the design and analysis of hydrogen-fueled gas-turbine combustion chambers.

The hydrodynamic theory of flame propagation in closed vessels is extended here to configurations involving nonuniform initial distributions of temperature and composition, admitting arbitrary steady-planar-deflagration chemical kinetics not considered in previous publications. The analysis addresses planar, cylindrical, and spherical configurations in the high-Péclet-number limit in which a thin premixed flame separates fresh-mixture and burnt-gas regions that are free from diffusive transport in the first asymptotic approximation. The simplified approach reduces the problem to solving a system of ordinary differential equations in time for the Euler-equation descriptions of the hydrodynamics in the fresh and burnt gases, along with the motion of the deflagration separating them. Planar hydrogen-air flames with a detailed chemical-kinetic description are selected to illustrate the simplified computational procedure and to expose the influences of the nonuniform initial distributions admitted by the new general formulation. The good accuracy of the simplified approach is supported by favorable agreement with a full numerical integration of the problem as originally formulated, prior to imposition of the asymptotic simplifications. The new method may facilitate future investigations of stratified-charge and unexpected-accident scenarios. Novelty and significance statement This work presents a reduced-order formulation of the evolution of flames in closed vessels, taking into account nonuniform initial distributions of temperature and reactant concentrations for the first time. It is significant in that it facilitates investigations of flame propagation under conditions that may arise in many real-world situations.

Melt-front instabilities during the combustion of a spinning polymethylmethacrylate disk in air are investigated. Mainly straight rivulet-type flow patterns were found, though under certain conditions saw-tooth patterns were observed. The measured wavelengths of the instabilities agree with earlier theoretical predictions of driven contact-line instabilities.

Starting with a detailed-chemistry description involving 20 elementary steps for hydrogen oxidation and 40 elementary steps for ammonia oxidation, it is shown that systematic application of sensitivity analyses of premixed flames under typical gas-turbine combustion conditions reduces the description to 12 elementary steps for hydrogen oxidation, 4 of them being reversible, and an additional 19 steps for ammonia oxidation, 6 of them being reversible, yielding reasonable predictions for auto-ignition and deflagration processes. Subsequent introduction of steady-state approximations for chemical intermediates, afforded by the high-pressure conditions existing in gas-turbine combustion chambers, effectively reduces the fuel-oxidation description in systems utilizing H 2 -NH 3 fuel mixtures to two global steps for deflagrations, namely, 2H 2 + O 2 ⇌ 2H 2 O and 4NH 3 ＋ 3O 2 ⇌ 2N 2 + 6 H 2 O. Analytical expressions for the associated overall rates, involving the local temperature and the O 2 , H 2 , NH 3 , N 2 , and H 2 O concentrations, are derived through selective truncation of the steady-state expressions, resulting in a simplified chemistry description that can facilitate future numerical analyses based on direct-numerical and large-eddy simulations. Novelty and significance statement A new short mechanism involving only 31 elementary reactions between 16 reactive species has been derived for hydrogen-ammonia oxidation under conditions of pressure, temperature and dilution typically found in gas-turbine burners. Introduction of steady-state assumptions for all intermediate species leads to a two-step mechanism that is shown to predict burning rates with sufficient accuracy. The proposed mechanism can significantly reduce computational times in future direct-numerical and large-eddy simulations.

Experimental and computational investigations are carried out to elucidate the influence of ethanol addition on n-heptane auto-ignition in counterflows. An axisymmetric stream of air, temperature gradually increased, is directed onto the surface of an evaporating pool of a liquid fuel. The air-stream temperature at auto-ignition is measured at various strain rates for n-heptane, ethanol, and various n-heptane/ethanol mixtures. Critical conditions for auto-ignition are predicted employing San Diego Mechanism for both fuels and fuel mixtures, and the results are compared with measurements. Measurements and predictions show that low-temperature chemistry (LTC) plays a significant role in promoting auto-ignition of n-heptane at low strain rates, but there is insufficient residence time at high strain rates for LTC to take place, so auto-ignition is promoted by high-temperature chemistry. Experimental and computational results show addition of ethanol inhibits LTC of n-heptane. To identify the responsible elementary steps, computations are performed to identify those dominate O2 consumption and contribute to the temperature rise in the reaction zone for n-heptane and n-heptane/ethanol mixtures at low strain rates. For n-heptane, O2 is consumed primarily by the low-temperature steps that result in ketohydroperoxide; the temperature rise is produced by subsequent LTC steps. For the mixtures, a key step consuming O2 is O2 + CH3CHOH = HO2 + CH3CHO, and the heat release occurs through the classical high-temperature reaction mechanism. Thus, the inhibition of auto-ignition that is observed to occur when ethanol is added to n-heptane arises from the competition for O2 between this step and the LTC addition of O2 to the heptyl radical and to the radical arising from the subsequent isomerization, for n-heptane.

Experimental and computational investigations are carried out to elucidate the influence of ethanol addition on n- heptane auto-ignition in counterflows. An axisymmetric stream of air, temperature gradually increased, is directed onto the surface of an evaporating pool of a liquid fuel. The air-stream temperature at auto-ignition is measured at various strain rates, defined as the axial gradient of the axial component of the flow velocity at the stagnation plane, for n- heptane, ethanol, and various n- heptane/ethanol mixtures. Critical conditions for auto-ignition are predicted employing the San Diego Mechanism for both fuels and the fuel mixtures, and the results are compared with the measurements. Measurements and predictions show that low-temperature chemistry plays a significant role in promoting auto-ignition of n- heptane at low strain rates, but there is insufficient residence time at high strain rates for low-temperature chemistry to take place, so auto-ignition is promoted by high-temperature chemistry. Experimental and computational results show that addition of ethanol inhibits the low-temperature chemistry of n- heptane. To identify the responsible elementary steps, computations are performed to identify those that dominate oxygen consumption and that contribute to the temperature rise in the reaction zone for n- heptane and n- heptane/ethanol mixtures at low strain rates. For n- heptane oxygen is consumed primarily by the low-temperature steps that result in ketohydroperoxide; the temperature rise is produced by subsequent low-temperature-chemistry steps. For the mixtures, a key step that consumes O 2 is O 2 + CH 3 CHOH = HO 2 + CH 3 CHO, and the heat release occurs through the classical high-temperature reaction mechanism. Thus, the inhibition of auto-ignition that is observed to occur when ethanol is added to n- heptane arises from the competition for oxygen between this step and the low-temperature-chemistry addition of O 2 to the heptyl radical and to the radical arising from the subsequent isomerization, for n- heptane.

Axisymmetric numerical simulations are used to assess the swirl-induced stabilization of low-Mach-number non-premixed jet flames at a moderate Reynolds number (Re=200). Using a one-step model chemistry describing methane-air partially premixed combustion, we carry out a parametric investigation of the coupling between vortex breakdown and laminar flame liftoff/blowoff in a concentric jet configuration involving a central non-swirling methane jet surrounded by a swirling annular air jet issuing from a pipe with radius RA′ rotating with angular speed Ω′. The analysis considers order-unity values of the two relevant controlling parameters, namely, the Damköhler number DN, defined as the square of the ratio of the stoichiometric methane-air flame-propagation velocity to the mean air-jet velocity UA′, and the swirl number S=Ω′RA′/UA′. As the Damköhler number DN is decreased the attached edge flame lifts off from the injector rim. The resulting lifted triple flame migrates downstream on further decreasing DN until a critical blowoff value DN,b is reached. Results for fixed S=1 exhibit lower values DN,b than the corresponding simulations with fixed S=0. For a fixed Damköhler number, it is found that increasing S results in increased entrainment and reduced liftoff heights. At a critical value SB* of the swirl number, equal to SB*=1.2 for DN=0.35, a recirculation zone abruptly forms upstream of the lifted triple flame, enhancing the mixing and facilitating flame stabilization closer to the injector.

The influence of adding a seventh important elementary step to a six-step mechanism, previously employed for describing the asymptotic structure of normal-alkane droplet combustion supported by cool-flame chemistry in the negative-temperature-coefficient (NTC) range, is investigated by analytical methods. A development paralleling the classical activation-energy-asymptotic (AEA) analysis of the partial-burning regime, accompanied for the first time by an AEA analysis for a negative activation energy, to account properly for the removal of an important intermediate species, is pursued to make predictions of the combustion process, resulting in a revised asymptotic structure that agrees better with computational predictions based on detailed chemistry.

Two-stage autoignition of n-dodecane droplets with varying ambient oxygen concentrations in oxygen-nitrogen mixtures are investigated experimentally under microgravity conditions using high-speed shadowgraphy. The ambient pressure and temperature are held constant at 3 atm and 650 K, respectively, while the droplet initial diameter is fixed approximately at 1.2 mm. During the two-stage autoignition process, first a cool-flame front forms in the leaner regions farther away from the evaporating droplet, and it then propagates toward the fuel-rich region closer to the droplet surface, eventually encompassing the droplet. A hot-flame kernel is then initiated in the wake of the cool flame and very quickly expands, establishing a classical diffusion flame around the droplet. The first and second ignition delay times are measured from shadowgraphic images captured at 3000 frames per second. The first induction time is found to be insensitive to the ambient oxygen concentration, while the second induction time varies approximately as the negative 2 power of the oxygen mole fraction.

A previously derived one-step reduced chemical-kinetic mechanism, describing hydrogen flames under near-limit conditions involving peak temperature not far from the crossover temperature, is used in computations of hydrogen-air flamelets at elevated pressures typical of gas-turbine combustion. Besides freely propagating laminar deflagrations with compositions spanning the whole range of flammability conditions, the calculations address strained premixed and nonpremixed flames as well as partially premixed propagating fronts. The comparisons with results of detailed-chemistry computations reveal that, for most purposes, the one-step mechanism provides sufficiently accurate predictions of burning rates under all conditions of interest for gas-turbine combustion. The reduced-chemistry model, featuring an explicit analytic expression for the hydrogen oxidation rate in terms of the local temperature and the O2, H2, and H2O concentrations, can be easily implemented in numerical codes, thereby facilitating future numerical analyses based on direct-numerical and large-eddy simulations.

The authors are a team of fire whirl researchers who have been actively studying whirls and large-scale wildland fires by directly observing them through fire-fighting efforts and applying theory, scale modeling, and numerical simulations in fire research. This multidisciplinary research-background team previously conducted scale model experiments to reconstruct hazardous large-scale fires in the laboratory, then conducted numerical simulations and developed fundamental theories to translate these findings into a basic understanding of combustion science and fluid dynamics. This article, a mix of reviews of the state of art experiments, theories, numerical modeling and artificial intelligence, and two case studies, is intended to address some safety concerns and raise awareness of large-scale fire whirls and forest fires with knowledge of thermodynamics, chemical kinetics, fluid dynamics, design, and practical fire-fighting experience, offering gaps that should be filled and future research to be conducted in each field, and crucial new observations and insights on large-scale fire incidents. We believe, this timely topic is of interest not only to fire research community but also to general readers, as the frequency and intensity of large-scale forest fires and fire whirls have increased, possibly due to the continuing global warming trend and human-induced changes in fuels. Each section and case study was written by one or two individual researchers based on their field of expertise which allows them to critically review progress made in their section of large-scale fire-whirls and forest-fires. Crucial observations and insights on the historical Great-Kanto-Earthquake-generated Hifukusho-Ato Fire-whirl (HAFW) and the slow rotations observed during recent forest firefighting efforts are presented. The first case study occurred in downtown Tokyo on 1 September 1923, as a result of the Great-Kanto-Earthquake, which claimed over 38,000 deaths within 15 min. The second case study discusses large-scale slow rotations observed during recent forest fires, which might had been responsible for the injuries and deaths of experienced firefighters.