The article Quasi-DNS Dataset of a Piloted Flame with Inhomogeneous Inlet Conditions written by T... more The article Quasi-DNS Dataset of a Piloted Flame with Inhomogeneous Inlet Conditions written by Thorsten Zirwes, Feichi Zhang, Peter Habisreuther, Maximilian Hansinger, Henning Bockhorn, Michael Pfitzner, Dimosthenis Trimis, was originally published Online First without Open Access.
Mass flow scaling of gas-assisted coaxial atomizers from laboratory to industrial scale is of maj... more Mass flow scaling of gas-assisted coaxial atomizers from laboratory to industrial scale is of major interest for a wide field of applications. However, there is only scarce knowledge and research concerning the effect of atomizer scale-up on liquid breakup and spray characteristics. The main objective of this study is therefore to derive basic principles for liquid jet breakup using upscaled nozzles to increase the liquid mass flow rate Ṁliq. For that purpose, atomizers with the same geometrical setup but increased sizes have been designed and experimentally investigated for Ṁliq=20, 50, 100, and 500 kg/h, while the aerodynamic Weber number Weaero and gas-to-liquid ratio GLR have been kept constant. The primary jet breakup was recorded via high-speed imaging, and the liquid core length LC and the frequency of the Kelvin–Helmholtz instability fK were extracted. Applying these results as reference data, highly resolved numerical simulations have been performed to gain a deeper under...
Fully resolved numerical simulations with finite rate chemical reactions and detailed molecular d... more Fully resolved numerical simulations with finite rate chemical reactions and detailed molecular diffusion have been conducted for a series of laminar premixed hydrogen–air flames under atmospheric condition. The objective of the work is to study the influence of unsteadiness of the flow on the local and global flame dynamics. Two equivalence ratios with Φ=0.5 and 4 are considered, leading to a negative and a positive Markstein number Ma0 at steady-state condition. The flames are excited with oscillating inflows at pre-defined frequencies f to assess the effect of unsteady flame stretch on flame dynamics. The Damköhler number, defined by the ratio of the inverse frequency of the oscillations and flame transit time, is used to characterize the interactions between the flow and the chemical reactions based on their time scales. For both lean and rich flame conditions, the local flame speed Sl is less sensitive to the flame stretch in an unsteady flow, which results in a reduced magnitu...
To meet future climate goals, the efficiency of combustion devices has to be increased. This requ... more To meet future climate goals, the efficiency of combustion devices has to be increased. This requires a better understanding of the underlying physics. The simulation of turbulent flames is a challenge because of the multi-scale nature of combustion processes: relevant length scales span over five orders of magnitude and time scales over more than ten. Because of this, the direct numerical simulation (DNS) of turbulent flames is only possible on large supercomputers. A DNS solver for chemically reacting flows implemented in the open-source framework Open-FOAM is presented. The thermo-chemical library Cantera is used to compute detailed transport coefficients based on kinetic gas theory. The multi-scale nature of time scales, which are much lower for the combustion chemistry than for the flow, are bridged by an operator splitting approach, which employs the open-source solver Sundials to integrate chemical reaction rates.
Highly-resolved numerical simulations have been performed for a coaxial, twin-fluid nozzle, which... more Highly-resolved numerical simulations have been performed for a coaxial, twin-fluid nozzle, which is designed for fundamental research of the atomization process in entrained flow gasification (EFG). Objective of the work is to reveal specifically the influence of the relative injection angle $\alpha$ between the central liquid jet and the annular airflow (external-mixing) on the primary atomization. A glycerol/water mixture with a dynamic viscosity of 200 mPa·s is atomized at atmospheric condition and with a gas-toliquid ratio of 0.6. The angle of attack between liquid and airflow has been varied from $\alpha$ = 0° to 60°. The computational grid consists of approx. 10 million cells, using a smallest resolution of 25 µm. In agreement with the corresponding experiment, the calculated breakup morphology is characterized by a pulsating mode destabilization, accompanied by disintegration of membrane-type ligaments. The breakup of the liquid jet is found to be enhanced from $\alpha$ = 0°...
Accurate quantification of heat release rate (HRR) profiles is useful for identifying important c... more Accurate quantification of heat release rate (HRR) profiles is useful for identifying important combustion processes in practical engines. This study aims to show whether the conventional HRR markers can still be used for highly stretched premixed methane/air flames with and without hydrogen addition. The correlation of formaldehyde-based molar concentrations, [H][CH 2 O] and [OH][CH 2 O], with HRR is numerically studied for highly stretched premixed CH 4 /air and CH 4 /H 2 /air flames. Two types of premixed flames are considered: the spherically expanding flame (SEF) starting from a highly positively stretched ignition kernel and the Bunsen flame with a high negative stretch rate at the flame tip. It is found that both [H][CH 2 O] and [OH][CH 2 O] can qualitatively describe the spatial distribution of HRR profiles in these two types of flames. In comparison to [H][CH 2 O], [OH][CH 2 O] is slightly better because it has a higher correlation coefficient and weaker sensitivity to the stoichiometry. However, during the ignition-influenced regime of SEFs, the magnitude of HRR cannot be accurately correlated by the concentrations of CH 2 O, OH, and H. Besides, hydrogen addition does not change the good spatial reconstruction qualities of these two HRR markers when its volume fraction in methane/hydrogen binary fuel blends is below 70%. The present results demonstrate that the conventional HRR markers, [OH][CH 2 O] and [H][CH 2 O], can be used to quantify the shape of HRR profiles even for highly stretched premixed flames with a small amount of hydrogen addition. Nevertheless, the magnitude of HRR cannot be accurately correlated by these two HRR markers for a highly stretched ignition kernel and hydrogen-dominated binary fuel blends. 48 on the total emission along the line of sight and is not spatially 49 resolved, 3 especially in complex flame topologies, such as 50 turbulent flames. 51 As an alternative, the LIF technique, which is spatially 52 resolved and can be extended to a planar imaging geometry 53 (PLIF), performs as a promising method to measure 54 concentrations of HRR indicators. The LIF measurement is 55 a null background absorption-based technique. It requires 56 sophisticated lasers to excite target species and then sample the 57 ground electronic state of the specific species. The formyl 58 radical, HCO, was first found to have a strong correlation with 59 HRR for laminar methane/air flames by Najm and his co
This work presents a numerical study on the effect of flame-wall interaction (FWI) from the viewp... more This work presents a numerical study on the effect of flame-wall interaction (FWI) from the viewpoint of flame dynamics. For that purpose, direct numerical simulations (DNS) employing detailed calculations of reaction rates and transport coefficients have been applied to a 2D premixed methane/air flame under atmospheric condition. Free flame (FF) and side-wall quenching (SWQ) configurations are realized by defining one lateral boundary as either a symmetry plane for the FF or a cold wall with fixed temperature at 20 o C for the SWQ case. Different components of flame stretch and Markstein number regarding tangential, normal (due to curvature) and total stretch, Ka s , Ka c and Ka tot = Ka s + Ka c , as well as their correlations with respect to the local flame consumption speed S L have been evaluated.
High Performance Computing in Science and Engineering ' 18, 2019
The computation of chemical reaction rates is commonly the performance bottleneck in CFD simulati... more The computation of chemical reaction rates is commonly the performance bottleneck in CFD simulations of turbulent combustion with detailed chemistry. Therefore, an optimization method is used where C++ source code is automatically generated for arbitrary reaction mechanisms. The generated code is highly optimized for the chosen mechanism and contains all routines for computing chemical reaction rates. In this work, the serial performance of the automatically generated source code, which in an earlier work only used ISO C++, is further improved by utilizing two compiler extensions: restrict and __builtin_assume_aligned. Introducing these two extensions to the generated code reduces the time for computing chemical reaction rates by up to 50% for the investigated reaction mechanism and total simulation time by up to 25%. Compared to OpenFOAM's standard chemistry implementation, the new code is faster by a factor of 10. This work discusses the effect of the two compiler extensions on performance by looking at two specific kernel functions from the automatically generated code and the effect on the assembly generated by the gcc and Intel compilers. The newly optimized code is used to evaluate the performance gain in a large scale parallel case, which simulates an experimentally investigated turbulent flame of laboratory scale on 14,400 CPU cores on the Hazel Hen cluster at HLRS. In the simulation, no combustion models are used and the flame is resolved down to the smallest length scales. With this approach, comparison of measured data with the simulation shows very good agreement. Using the optimized code including compiler extensions, total simulation time decreases by 20% compared to the same code without compiler extensions. A comprehensive database from the simulation results has been assembled and will consist of 10 TB of 3D and 2D transient field variables.
Mechanical sparks play an important role in safety-related assessments of the ignition of combust... more Mechanical sparks play an important role in safety-related assessments of the ignition of combustible mixtures in process industry and aviation. They are mechanically generated from solid materials by grinding, abrasive cutting, or mechanical impact. In general, the particle temperature required for ignition sharply increases with decreasing particle size [1–6]. Another important factor is the relative velocity between particle and gas phase. Using electrically heated cylindrical rods, Mullen et al. [7] studied the ignition of different fuel/air mixtures at various fluid velocities between 25 and 150 m/s. Higher ignition temperature was recorded for increasing velocity, lower rod diameters as well as lower initial gas temperature and pressure. High-speed photography showed that ignition begins in the stagnant region near the downstream face of the heated cylinder. The results are consistent with earlier experiments by Paterson [8] who shot hot 2 mm diameter platinum and 3.5 mm diame...
Flame particles (FP) are massless, virtual particles which follow material points on the flame su... more Flame particles (FP) are massless, virtual particles which follow material points on the flame surface. This work presents a tracking algorithm for FPs which utilizes barycentric coordinates. The methodology can be used with any cell shape in the computational mesh and allows computationally fast spatial interpolation as well as efficient determination of the intersection of FP trajectories with iso-surfaces. In contrast to previous flame particle tracking (FPT) approaches, the code is fully parallelized and can therefore be used in-situ during the simulation. It also includes fully parallelized computation of flame consumption speed by integrating reaction rates along a line normal to the flame surface at each FP position. Direct numerical simulations of laminar pulsating premixed hydrogen-air Bunsen flames serve as validation cases and showcase the added value of tracking material points for studying local flame dynamics. Exciting the inlet flow harmonically with frequencies equal to the inverse flame time scale leads to a pulsating mode where the flame front is corrugated. Ten times higher frequencies nearly resemble the steady state solution. The FPs are seeded along the flame surface and are used to track the unsteady diffusive, convective and chemical contributions at arbitrary points on the flame front over time. Their trajectories reveal a phase shift between the unsteady flame stretch rate and local flame speed of the order of 0.1 flame time scales for rich hydrogen flames. This is caused by a time delay between straining and stretch due to curvature. The reason is that diffusive processes follow the time signal of curvature while chemical processes are most strongly affected by the straining rate, which
Highly-resolved numerical simulations employing detailed reaction kinetics and molecular transpor... more Highly-resolved numerical simulations employing detailed reaction kinetics and molecular transport have been applied to flame-wall interaction (FWI) of laminar premixed flames. A multiple plane-jet flame (2D) has been considered, which is operated with premixed methane/air mixtures at atmospheric conditions and with different equivalence ratios. Free flame (FF) and side-wall quenching (SWQ) conditions have been accomplished by defining one lateral boundary as either a symmetry plane for FF or a cold wall with fixed temperature for SWQ. An equidistant grid with a resolution of 20 µm is used to resolve the FWI zone. The GRI-3.0 mechanism is used for computing chemical reaction rates. The flame is tangentially compressed when approaching the cold wall, and elongated in the FF case, causing an inversion of the sign of the tangential strain rate Ka s and a considerable decrease of the total stretch rate Ka tot for the SWQ flame. The flame consumption speed S L decreases with decreasing normal stretch due to curvature Ka c while approaching the cold wall, but it increases with decreasing Ka c for the FF case, leading to an inversion of the Markstein number Ma tot based on Ka tot from positive in FF to negative in the SWQ case. The results reveal a strong correlation of flame dynamics during transitions from FWI to freely propagating flames, which may bring a new perspective for modeling FWI phenomena by means of flame dynamics. To do this, the quenching effect of the wall may be reproduced by an inversion of the Markstein number from positive to negative in the FWI zone and applying the general linear Markstein correlation, leading to a decrease of the flame consumption speed. In addition, the quenching distance evaluated from S L has been found to be almost equal to the unstretched laminar flame thickness, which compares quantitatively well with measured data from literature.
The numerical investigation of quenching distances in laminar flows is mainly concerned with two ... more The numerical investigation of quenching distances in laminar flows is mainly concerned with two setups: head-on quenching (HOQ) and side-wall quenching (SWQ). While most of the numerical work has been conducted for HOQ with good agreement between simulation and experiment, far less analysis has been done for SWQ. Most of the SWQ simulations used simplified diffusion models or reduced chemistry and achieved reasonable agreement with experiments. However, it has been found that quenching distances for the SWQ setup differ from experimental results if detailed diffusion models and chemical reaction mechanisms are employed. Side-wall quenching is investigated numerically in this work with steady-state 2D and 3D simulations of an experimental flame setup. The simulations fully resolve the flame and employ detailed reaction mechanisms as well as molecular diffusion models. The goal is to provide data for the sensitivity of numerical quenching distances to different parameters. Quenching ...
Understanding and characterizing ignition of flammable mixtures by hot particles is important for... more Understanding and characterizing ignition of flammable mixtures by hot particles is important for assessing and reducing the risk of accidental ignition and explosion in industry and aviation. Recently, many studies have been conducted for ignition of gaseous mixtures by hot particles. However, the effects of lowtemperature chemistry (LTC) on ignition by hot particles received little attention. LTC plays an important role in the ignition of most hydrocarbon fuels and may induce cool flames. The present study aims to numerically assess the effects of LTC on ignition by the hot particles. We consider the transient ignition processes induced by a hot spherical particle in quiescent and flowing stoichiometric dimethyl ether/air mixtures. 1D and 2D simulations, respectively, are conducted for the ignition process by hot-particles in quiescent and flowing mixtures. A detailed kinetic model including both LTC and high-temperature chemistry (HTC) is used in simulations. The results exhibit a premixed cool flame to be first initiated by the hot particle. Then a double-flame structure with both premixed cool and hot flames is observed at certain conditions. At zero or low inlet flow velocities, the hot flame catches up and merges with the leading cool flame. At high inlet flow velocities, the hot flame cannot be initiated due to the short residence time and large convective loss of heat and radicals. Comparing the results with and without considering LTC confirms that LTC accelerates substantially ignition via HTC in a certain range of hot particle temperatures. The mechanism of ignition promotion by LTC is interpreted by analyzing the radical pool produced by the LTC and HTC surrounding the hot particle. Moreover, the influence of inlet flow velocity on ignition by hot particles is assessed. Non-monotonic change of ignition delay time with flow velocity is observed and discussed.
Identifying combustion regimes in terms of premixed and non-premixed characteristics is an import... more Identifying combustion regimes in terms of premixed and non-premixed characteristics is an important task for understanding combustion phenomena and the structure of flames. A quasi-DNS database of the compositionally inhomogeneous partially premixed Sydney/Sandia flame in configuration FJ-5GP-Lr75-57 is used to directly compare different types of flame regime markers from literature. In the simulation of the flame, detailed chemistry and diffusion models are utilized and no turbulence and combustion models are used as the flame front and flow are fully resolved near the nozzle. This allows evaluating the regime markers as a post-processing step without modeling assumptions and directly comparing regime markers based on gradient alignment, drift term analysis and gradient free regime identification. The goal is not to find the correct regime marker, which might be impossible due to the different set of assumptions of every marker and the generally vague definition of the partially p...
Simulations of primary breakup process of high-viscous liquid jet at elevated pressures • Detaile... more Simulations of primary breakup process of high-viscous liquid jet at elevated pressures • Detailed resolution of near-field multiphase interactions for elucidating the breakup mechanism • Effect of elevated pressure is equivalent to that of increased momentum flux ratio • Strongly enhanced multiphase momentum exchange at elevated pressure • First-order estimate proposed for assessing spray dynamics/droplet-velocity
Large Eddy Simulations of the Sydney mixed-mode flame with inhomogeneous inlet (FJ200-5GP-Lr75-57... more Large Eddy Simulations of the Sydney mixed-mode flame with inhomogeneous inlet (FJ200-5GP-Lr75-57) are performed using the Eulerian Stochastic Fields (ESF) transported probability functions method to account for the sub-grid scale turbulence–chemistry interaction, to demonstrate the suitability of the ESF method for mixed-mode combustion. An analytically reduced 19-species methane mechanism is used for the description of the chemical reactions. Prior to the reactive case, simulation results of the non-reactive setup with cold and hot pilot stream are presented, which show differences in the jet breakup and radial species mass fluxes. The reactive case simulations are compared to experimental data and a recently conducted model free quasi-DNS (qDNS), showing very good agreement with the qDNS in terms of scatter data and radial mean values of temperature and species distribution, as well as mixture fraction conditional statistics. Further analysis is dedicated to sub-grid scale statis...
A quasi-DNS of the partially premixed turbulent Sydney flame in configuration FJ200-5GP-Lr75-57 h... more A quasi-DNS of the partially premixed turbulent Sydney flame in configuration FJ200-5GP-Lr75-57 has been conducted using detailed molecular diffusion for multi-component mixtures and complex reaction mechanisms. In order to study flame dynamics like regime transition in this flame for the development of new combustion models and to directly compare the quasi-DNS to different LES models, the simulation results are compiled into a data base. Because the simulation was performed with OpenFOAM, we demonstrate the quasi-DNS capabilities of OpenFOAM by performing canonical test cases. They attest that OpenFOAM’s cubic discretization has lower numerical diffusion compared to classical central difference schemes and can reach higher than second order convergence rate in some cases. The quasi-DNS of the Sydney flame is conducted with a self-developed reacting flow solver which is able to accurately compute molecular diffusion coefficients from kinetic gas theory and employs a fast implementa...
The article Quasi-DNS Dataset of a Piloted Flame with Inhomogeneous Inlet Conditions written by T... more The article Quasi-DNS Dataset of a Piloted Flame with Inhomogeneous Inlet Conditions written by Thorsten Zirwes, Feichi Zhang, Peter Habisreuther, Maximilian Hansinger, Henning Bockhorn, Michael Pfitzner, Dimosthenis Trimis, was originally published Online First without Open Access.
Mass flow scaling of gas-assisted coaxial atomizers from laboratory to industrial scale is of maj... more Mass flow scaling of gas-assisted coaxial atomizers from laboratory to industrial scale is of major interest for a wide field of applications. However, there is only scarce knowledge and research concerning the effect of atomizer scale-up on liquid breakup and spray characteristics. The main objective of this study is therefore to derive basic principles for liquid jet breakup using upscaled nozzles to increase the liquid mass flow rate Ṁliq. For that purpose, atomizers with the same geometrical setup but increased sizes have been designed and experimentally investigated for Ṁliq=20, 50, 100, and 500 kg/h, while the aerodynamic Weber number Weaero and gas-to-liquid ratio GLR have been kept constant. The primary jet breakup was recorded via high-speed imaging, and the liquid core length LC and the frequency of the Kelvin–Helmholtz instability fK were extracted. Applying these results as reference data, highly resolved numerical simulations have been performed to gain a deeper under...
Fully resolved numerical simulations with finite rate chemical reactions and detailed molecular d... more Fully resolved numerical simulations with finite rate chemical reactions and detailed molecular diffusion have been conducted for a series of laminar premixed hydrogen–air flames under atmospheric condition. The objective of the work is to study the influence of unsteadiness of the flow on the local and global flame dynamics. Two equivalence ratios with Φ=0.5 and 4 are considered, leading to a negative and a positive Markstein number Ma0 at steady-state condition. The flames are excited with oscillating inflows at pre-defined frequencies f to assess the effect of unsteady flame stretch on flame dynamics. The Damköhler number, defined by the ratio of the inverse frequency of the oscillations and flame transit time, is used to characterize the interactions between the flow and the chemical reactions based on their time scales. For both lean and rich flame conditions, the local flame speed Sl is less sensitive to the flame stretch in an unsteady flow, which results in a reduced magnitu...
To meet future climate goals, the efficiency of combustion devices has to be increased. This requ... more To meet future climate goals, the efficiency of combustion devices has to be increased. This requires a better understanding of the underlying physics. The simulation of turbulent flames is a challenge because of the multi-scale nature of combustion processes: relevant length scales span over five orders of magnitude and time scales over more than ten. Because of this, the direct numerical simulation (DNS) of turbulent flames is only possible on large supercomputers. A DNS solver for chemically reacting flows implemented in the open-source framework Open-FOAM is presented. The thermo-chemical library Cantera is used to compute detailed transport coefficients based on kinetic gas theory. The multi-scale nature of time scales, which are much lower for the combustion chemistry than for the flow, are bridged by an operator splitting approach, which employs the open-source solver Sundials to integrate chemical reaction rates.
Highly-resolved numerical simulations have been performed for a coaxial, twin-fluid nozzle, which... more Highly-resolved numerical simulations have been performed for a coaxial, twin-fluid nozzle, which is designed for fundamental research of the atomization process in entrained flow gasification (EFG). Objective of the work is to reveal specifically the influence of the relative injection angle $\alpha$ between the central liquid jet and the annular airflow (external-mixing) on the primary atomization. A glycerol/water mixture with a dynamic viscosity of 200 mPa·s is atomized at atmospheric condition and with a gas-toliquid ratio of 0.6. The angle of attack between liquid and airflow has been varied from $\alpha$ = 0° to 60°. The computational grid consists of approx. 10 million cells, using a smallest resolution of 25 µm. In agreement with the corresponding experiment, the calculated breakup morphology is characterized by a pulsating mode destabilization, accompanied by disintegration of membrane-type ligaments. The breakup of the liquid jet is found to be enhanced from $\alpha$ = 0°...
Accurate quantification of heat release rate (HRR) profiles is useful for identifying important c... more Accurate quantification of heat release rate (HRR) profiles is useful for identifying important combustion processes in practical engines. This study aims to show whether the conventional HRR markers can still be used for highly stretched premixed methane/air flames with and without hydrogen addition. The correlation of formaldehyde-based molar concentrations, [H][CH 2 O] and [OH][CH 2 O], with HRR is numerically studied for highly stretched premixed CH 4 /air and CH 4 /H 2 /air flames. Two types of premixed flames are considered: the spherically expanding flame (SEF) starting from a highly positively stretched ignition kernel and the Bunsen flame with a high negative stretch rate at the flame tip. It is found that both [H][CH 2 O] and [OH][CH 2 O] can qualitatively describe the spatial distribution of HRR profiles in these two types of flames. In comparison to [H][CH 2 O], [OH][CH 2 O] is slightly better because it has a higher correlation coefficient and weaker sensitivity to the stoichiometry. However, during the ignition-influenced regime of SEFs, the magnitude of HRR cannot be accurately correlated by the concentrations of CH 2 O, OH, and H. Besides, hydrogen addition does not change the good spatial reconstruction qualities of these two HRR markers when its volume fraction in methane/hydrogen binary fuel blends is below 70%. The present results demonstrate that the conventional HRR markers, [OH][CH 2 O] and [H][CH 2 O], can be used to quantify the shape of HRR profiles even for highly stretched premixed flames with a small amount of hydrogen addition. Nevertheless, the magnitude of HRR cannot be accurately correlated by these two HRR markers for a highly stretched ignition kernel and hydrogen-dominated binary fuel blends. 48 on the total emission along the line of sight and is not spatially 49 resolved, 3 especially in complex flame topologies, such as 50 turbulent flames. 51 As an alternative, the LIF technique, which is spatially 52 resolved and can be extended to a planar imaging geometry 53 (PLIF), performs as a promising method to measure 54 concentrations of HRR indicators. The LIF measurement is 55 a null background absorption-based technique. It requires 56 sophisticated lasers to excite target species and then sample the 57 ground electronic state of the specific species. The formyl 58 radical, HCO, was first found to have a strong correlation with 59 HRR for laminar methane/air flames by Najm and his co
This work presents a numerical study on the effect of flame-wall interaction (FWI) from the viewp... more This work presents a numerical study on the effect of flame-wall interaction (FWI) from the viewpoint of flame dynamics. For that purpose, direct numerical simulations (DNS) employing detailed calculations of reaction rates and transport coefficients have been applied to a 2D premixed methane/air flame under atmospheric condition. Free flame (FF) and side-wall quenching (SWQ) configurations are realized by defining one lateral boundary as either a symmetry plane for the FF or a cold wall with fixed temperature at 20 o C for the SWQ case. Different components of flame stretch and Markstein number regarding tangential, normal (due to curvature) and total stretch, Ka s , Ka c and Ka tot = Ka s + Ka c , as well as their correlations with respect to the local flame consumption speed S L have been evaluated.
High Performance Computing in Science and Engineering ' 18, 2019
The computation of chemical reaction rates is commonly the performance bottleneck in CFD simulati... more The computation of chemical reaction rates is commonly the performance bottleneck in CFD simulations of turbulent combustion with detailed chemistry. Therefore, an optimization method is used where C++ source code is automatically generated for arbitrary reaction mechanisms. The generated code is highly optimized for the chosen mechanism and contains all routines for computing chemical reaction rates. In this work, the serial performance of the automatically generated source code, which in an earlier work only used ISO C++, is further improved by utilizing two compiler extensions: restrict and __builtin_assume_aligned. Introducing these two extensions to the generated code reduces the time for computing chemical reaction rates by up to 50% for the investigated reaction mechanism and total simulation time by up to 25%. Compared to OpenFOAM's standard chemistry implementation, the new code is faster by a factor of 10. This work discusses the effect of the two compiler extensions on performance by looking at two specific kernel functions from the automatically generated code and the effect on the assembly generated by the gcc and Intel compilers. The newly optimized code is used to evaluate the performance gain in a large scale parallel case, which simulates an experimentally investigated turbulent flame of laboratory scale on 14,400 CPU cores on the Hazel Hen cluster at HLRS. In the simulation, no combustion models are used and the flame is resolved down to the smallest length scales. With this approach, comparison of measured data with the simulation shows very good agreement. Using the optimized code including compiler extensions, total simulation time decreases by 20% compared to the same code without compiler extensions. A comprehensive database from the simulation results has been assembled and will consist of 10 TB of 3D and 2D transient field variables.
Mechanical sparks play an important role in safety-related assessments of the ignition of combust... more Mechanical sparks play an important role in safety-related assessments of the ignition of combustible mixtures in process industry and aviation. They are mechanically generated from solid materials by grinding, abrasive cutting, or mechanical impact. In general, the particle temperature required for ignition sharply increases with decreasing particle size [1–6]. Another important factor is the relative velocity between particle and gas phase. Using electrically heated cylindrical rods, Mullen et al. [7] studied the ignition of different fuel/air mixtures at various fluid velocities between 25 and 150 m/s. Higher ignition temperature was recorded for increasing velocity, lower rod diameters as well as lower initial gas temperature and pressure. High-speed photography showed that ignition begins in the stagnant region near the downstream face of the heated cylinder. The results are consistent with earlier experiments by Paterson [8] who shot hot 2 mm diameter platinum and 3.5 mm diame...
Flame particles (FP) are massless, virtual particles which follow material points on the flame su... more Flame particles (FP) are massless, virtual particles which follow material points on the flame surface. This work presents a tracking algorithm for FPs which utilizes barycentric coordinates. The methodology can be used with any cell shape in the computational mesh and allows computationally fast spatial interpolation as well as efficient determination of the intersection of FP trajectories with iso-surfaces. In contrast to previous flame particle tracking (FPT) approaches, the code is fully parallelized and can therefore be used in-situ during the simulation. It also includes fully parallelized computation of flame consumption speed by integrating reaction rates along a line normal to the flame surface at each FP position. Direct numerical simulations of laminar pulsating premixed hydrogen-air Bunsen flames serve as validation cases and showcase the added value of tracking material points for studying local flame dynamics. Exciting the inlet flow harmonically with frequencies equal to the inverse flame time scale leads to a pulsating mode where the flame front is corrugated. Ten times higher frequencies nearly resemble the steady state solution. The FPs are seeded along the flame surface and are used to track the unsteady diffusive, convective and chemical contributions at arbitrary points on the flame front over time. Their trajectories reveal a phase shift between the unsteady flame stretch rate and local flame speed of the order of 0.1 flame time scales for rich hydrogen flames. This is caused by a time delay between straining and stretch due to curvature. The reason is that diffusive processes follow the time signal of curvature while chemical processes are most strongly affected by the straining rate, which
Highly-resolved numerical simulations employing detailed reaction kinetics and molecular transpor... more Highly-resolved numerical simulations employing detailed reaction kinetics and molecular transport have been applied to flame-wall interaction (FWI) of laminar premixed flames. A multiple plane-jet flame (2D) has been considered, which is operated with premixed methane/air mixtures at atmospheric conditions and with different equivalence ratios. Free flame (FF) and side-wall quenching (SWQ) conditions have been accomplished by defining one lateral boundary as either a symmetry plane for FF or a cold wall with fixed temperature for SWQ. An equidistant grid with a resolution of 20 µm is used to resolve the FWI zone. The GRI-3.0 mechanism is used for computing chemical reaction rates. The flame is tangentially compressed when approaching the cold wall, and elongated in the FF case, causing an inversion of the sign of the tangential strain rate Ka s and a considerable decrease of the total stretch rate Ka tot for the SWQ flame. The flame consumption speed S L decreases with decreasing normal stretch due to curvature Ka c while approaching the cold wall, but it increases with decreasing Ka c for the FF case, leading to an inversion of the Markstein number Ma tot based on Ka tot from positive in FF to negative in the SWQ case. The results reveal a strong correlation of flame dynamics during transitions from FWI to freely propagating flames, which may bring a new perspective for modeling FWI phenomena by means of flame dynamics. To do this, the quenching effect of the wall may be reproduced by an inversion of the Markstein number from positive to negative in the FWI zone and applying the general linear Markstein correlation, leading to a decrease of the flame consumption speed. In addition, the quenching distance evaluated from S L has been found to be almost equal to the unstretched laminar flame thickness, which compares quantitatively well with measured data from literature.
The numerical investigation of quenching distances in laminar flows is mainly concerned with two ... more The numerical investigation of quenching distances in laminar flows is mainly concerned with two setups: head-on quenching (HOQ) and side-wall quenching (SWQ). While most of the numerical work has been conducted for HOQ with good agreement between simulation and experiment, far less analysis has been done for SWQ. Most of the SWQ simulations used simplified diffusion models or reduced chemistry and achieved reasonable agreement with experiments. However, it has been found that quenching distances for the SWQ setup differ from experimental results if detailed diffusion models and chemical reaction mechanisms are employed. Side-wall quenching is investigated numerically in this work with steady-state 2D and 3D simulations of an experimental flame setup. The simulations fully resolve the flame and employ detailed reaction mechanisms as well as molecular diffusion models. The goal is to provide data for the sensitivity of numerical quenching distances to different parameters. Quenching ...
Understanding and characterizing ignition of flammable mixtures by hot particles is important for... more Understanding and characterizing ignition of flammable mixtures by hot particles is important for assessing and reducing the risk of accidental ignition and explosion in industry and aviation. Recently, many studies have been conducted for ignition of gaseous mixtures by hot particles. However, the effects of lowtemperature chemistry (LTC) on ignition by hot particles received little attention. LTC plays an important role in the ignition of most hydrocarbon fuels and may induce cool flames. The present study aims to numerically assess the effects of LTC on ignition by the hot particles. We consider the transient ignition processes induced by a hot spherical particle in quiescent and flowing stoichiometric dimethyl ether/air mixtures. 1D and 2D simulations, respectively, are conducted for the ignition process by hot-particles in quiescent and flowing mixtures. A detailed kinetic model including both LTC and high-temperature chemistry (HTC) is used in simulations. The results exhibit a premixed cool flame to be first initiated by the hot particle. Then a double-flame structure with both premixed cool and hot flames is observed at certain conditions. At zero or low inlet flow velocities, the hot flame catches up and merges with the leading cool flame. At high inlet flow velocities, the hot flame cannot be initiated due to the short residence time and large convective loss of heat and radicals. Comparing the results with and without considering LTC confirms that LTC accelerates substantially ignition via HTC in a certain range of hot particle temperatures. The mechanism of ignition promotion by LTC is interpreted by analyzing the radical pool produced by the LTC and HTC surrounding the hot particle. Moreover, the influence of inlet flow velocity on ignition by hot particles is assessed. Non-monotonic change of ignition delay time with flow velocity is observed and discussed.
Identifying combustion regimes in terms of premixed and non-premixed characteristics is an import... more Identifying combustion regimes in terms of premixed and non-premixed characteristics is an important task for understanding combustion phenomena and the structure of flames. A quasi-DNS database of the compositionally inhomogeneous partially premixed Sydney/Sandia flame in configuration FJ-5GP-Lr75-57 is used to directly compare different types of flame regime markers from literature. In the simulation of the flame, detailed chemistry and diffusion models are utilized and no turbulence and combustion models are used as the flame front and flow are fully resolved near the nozzle. This allows evaluating the regime markers as a post-processing step without modeling assumptions and directly comparing regime markers based on gradient alignment, drift term analysis and gradient free regime identification. The goal is not to find the correct regime marker, which might be impossible due to the different set of assumptions of every marker and the generally vague definition of the partially p...
Simulations of primary breakup process of high-viscous liquid jet at elevated pressures • Detaile... more Simulations of primary breakup process of high-viscous liquid jet at elevated pressures • Detailed resolution of near-field multiphase interactions for elucidating the breakup mechanism • Effect of elevated pressure is equivalent to that of increased momentum flux ratio • Strongly enhanced multiphase momentum exchange at elevated pressure • First-order estimate proposed for assessing spray dynamics/droplet-velocity
Large Eddy Simulations of the Sydney mixed-mode flame with inhomogeneous inlet (FJ200-5GP-Lr75-57... more Large Eddy Simulations of the Sydney mixed-mode flame with inhomogeneous inlet (FJ200-5GP-Lr75-57) are performed using the Eulerian Stochastic Fields (ESF) transported probability functions method to account for the sub-grid scale turbulence–chemistry interaction, to demonstrate the suitability of the ESF method for mixed-mode combustion. An analytically reduced 19-species methane mechanism is used for the description of the chemical reactions. Prior to the reactive case, simulation results of the non-reactive setup with cold and hot pilot stream are presented, which show differences in the jet breakup and radial species mass fluxes. The reactive case simulations are compared to experimental data and a recently conducted model free quasi-DNS (qDNS), showing very good agreement with the qDNS in terms of scatter data and radial mean values of temperature and species distribution, as well as mixture fraction conditional statistics. Further analysis is dedicated to sub-grid scale statis...
A quasi-DNS of the partially premixed turbulent Sydney flame in configuration FJ200-5GP-Lr75-57 h... more A quasi-DNS of the partially premixed turbulent Sydney flame in configuration FJ200-5GP-Lr75-57 has been conducted using detailed molecular diffusion for multi-component mixtures and complex reaction mechanisms. In order to study flame dynamics like regime transition in this flame for the development of new combustion models and to directly compare the quasi-DNS to different LES models, the simulation results are compiled into a data base. Because the simulation was performed with OpenFOAM, we demonstrate the quasi-DNS capabilities of OpenFOAM by performing canonical test cases. They attest that OpenFOAM’s cubic discretization has lower numerical diffusion compared to classical central difference schemes and can reach higher than second order convergence rate in some cases. The quasi-DNS of the Sydney flame is conducted with a self-developed reacting flow solver which is able to accurately compute molecular diffusion coefficients from kinetic gas theory and employs a fast implementa...
Uploads
Papers by Feichi Zhang