INJECTOR DESIGN SPACE EXPLORATION FOR AN ULTRA-LOW NO X HYDROGEN MICROMIX COMBUSTION SYSTEM

The depletion of fossil fuel resources, as well as the increasing environmental concerns have become the driving forces towards the research and development necessary for the introduction of alternative fuel such as hydrogen into civil aviation. Hydrogen is a suitable energy source primarily because it is free of carbon and other forms of impurities and is also the most abundant element in the universe. The advantages of using Liquid Hydrogen (LH 2 ) for civil aviation extends beyond carbon-free mission level emissions; LH 2 combustion can potentially reduce NO x emission by up to 90%, providing long-term sustainability and unrivalled environmental benefits. The paper presents a simplified parametric analysis to investigate the influence of various injector design parameters on a hydrogen micromix combustor reactive flow field. The main characteristics investigated are the flame structure (shape and position), the aerodynamic stabilization of the flame and the resulting NO x emissions. The design parameters include variations in the air-feed dimensions and the hydrogen injection diameter. A suitable numerical model was established by comparing various turbulence modelling approaches, reaction mechanisms and turbulence-chemistry interaction modelling schemes. The predictive capabilities, and limitations, of each of these modelling approaches, are assessed. The numerical challenges and limitations associated with modelling H 2 /air combustion at high pressure and temperature conditions are detailed. The influence of varying the injector design parameters on the mixing and hence the NO x characteristics is assessed.


ABSTRACT
The depletion of fossil fuel resources, as well as the increasing environmental concerns have become the driving forces towards the research and development necessary for the introduction of alternative fuel such as hydrogen into civil aviation.Hydrogen is a suitable energy source primarily because it is free of carbon and other forms of impurities and is also the most abundant element in the universe.The advantages of using Liquid Hydrogen (LH2) for civil aviation extends beyond carbon-free mission level emissions; LH2 combustion can potentially reduce NOx emission by up to 90%, providing long-term sustainability and unrivalled environmental benefits.
The paper presents a simplified parametric analysis to investigate the influence of various injector design parameters on a hydrogen micromix combustor reactive flow field.The main characteristics investigated are the flame structure (shape and position), the aerodynamic stabilization of the flame and the resulting NOx emissions.The design parameters include variations in the air-feed dimensions and the hydrogen injection diameter.A suitable numerical model was established by comparing various turbulence modelling approaches, reaction mechanisms and turbulence-chemistry interaction modelling schemes.The predictive capabilities, and limitations, of each of these modelling approaches, are assessed.The numerical challenges and limitations associated with modelling H2/air combustion at high pressure and temperature conditions are detailed.The influence of varying the injector design parameters on the mixing and hence the NOx characteristics is assessed.

INTRODUCTION
The continued growth of the aviation industry, combined with growing environmental concerns and the depletion of the fossil fuel resources makes it is necessary to develop clean and efficient technologies for a sustainable growth in aviation.The total consumption of aviation fuel was estimated to be 6.8 million barrels a day in 2007 [1].Currently, the aviation fuel constitutes about 6% of the oil consumption globally.This accounts for about 2.1% of total anthropogenic CO2 emissions or 12% of the total carbon emissions from all transport sources [2].Aircraft NOx emissions contribute about 75-80% of the total airport NOx emissions.These figures are for the year 2016, however, IATA projects an increase in air traffic by about 3-5% a year for the passenger transport sector and about 9-10% for the cargo sector over the next 20 years [3].
To curb the environmental impact of the aviation industry, ambitious targets have been set by flightpath 2050 which includes 75% reduction in CO2 emission per passage kilometre, 90% reduction in NOx emissions and 65% reduction in perceived noise emission relative to the year 2000 technologies [4].However, increasing evidence suggests that these targets would be very challenging to meet with conventional carbon-containing fuels, despite the large research efforts ongoing for airframe, propulsion technologies and improved asset management.LH2 as a fuel provides an alternative to meet these highly ambitious targets.
The emissions resulting from LH2-fueled combustion systems differ significantly from hydrocarbon combustion.LH2 combustion does not produce any carbon pollutant species, in fact, the primary combustion products from hydrogen combustion is the formattion of water vapour.Water vapour is also considered a greenhouse emission however, unlike hydrocarbons, its effect depends upon the altitude of the cirrus cloud formation.Moreover, the residence time of water vapour is relatively short; several days or up to one year [5] [6].
The thermodynamic cycle of a gas turbine engine demands that he combustion processes take place at relatively high levels of temperature, which favour the formation of NOx.The combustion of LH2 as a fuel has a greater potential to generate higher NOx than conventional kerosene fuel primarily because of its higher flame temperature at a given equivalence ratio (Φ).However, since LH2 has a wider flammability limit, the entire combustion zone can be made significantly leaner, resulting in lower flame temperatures and hence reduced NOx emissions.Figure 1 presents the flame temperature for hydrogen and kerosene as a function of the combustor primary zone Φ.
Premixing of an LH2 combustor would, therefore, be very desirable from the low-NOx prospective.However, obtaining primary zone Φ uniformly low enough to inhibit excessive NOx values is quite challenging; imperfections in the fuel/air mixing can potentially lead to local rich fuel-air pockets, leading to a greater presence of hot spots and corresponding peaks in NOx production.Another issue that can potentially arise with a premixed system is the enhanced risk of autoignition and flashback due to the high reactivity of LH2-air mixtures [8].The high stoichiometric temperature, and the high reactivity associated with hydrogen, therefore makes it imperative to have the necessary methods and tools available for the design fuel injection systems that can ensure good mixture homogeneity while, at the same time, mitigating the risks associated with autoignition and flashback.

Figure 1-Temperature Characteristics [7]
Although the uptake of LH2 for aviation is still in its infancy, it can be assumed that the LH2 will play a vital role in alleviating environmental impact of the aviation.It is therefore vital to develop both the analytical and experimental tools to further mature the technologies to higher TRL levels for the use of LH2 for civil aviation.The micromix combustion principle has been investigated at Cranfield University over the past years [9][10][11][12][13][14][15].The scope of the current work will feed into the research and innovation efforts by Cranfield University as a part of the ENABLEH2 project [19].This involves the design and development of a novel ultra-low-NOx micromix combustor system to TRL levels 3-4.
The present study outlines the preliminary numerical methodology adopted to explore the design space and the reactive flow fields associated with micromix combustion.The study has been subdivided into two parts.The first part involves a turbulence-chemistry interaction modelling study using three different reduced hydrogen chemistry mechanisms, at high pressure and temperature condition.The applicability and predictive capabilities of each of these mechanisms are analysed to down-select the best candidate for further parametric analysis.The second part of this study involves a preliminary parametric analysis.As a part of this parametric study the influence of varying the air feed hydraulic diameter and the hydrogen inlet diameter on the reactive flow field, and hence the flame aerodynamic stability is analysed.

Micromix Combustion Concept
The combustion characteristics of hydrogen differ significantly from that of kerosene as stated above.Therefore, to ensure good mixing homogeneity and to avoid issues associated with flashback and autoignition, the concept of "micromix combustion" was conceived by the FHA Aachen University [8].The micromix combustion concept can be characterised as an method to, "minimise the scale of mixing but to increase the mixing intensity" [8].The mixing of the fuel and air in the local zone is significantly increased compared to conventional gas turbine combustor designs.This maximising of the mixing intensity is however limited by the maximum pressure loss of modern gas turbine combustor designs, which should not typically exceed 3-4% [8].
The mixing process in a micromix system is significantly improved by the principle of the jet-in-cross-flow.The flame anchoring, which is essential for the reduction in NOx formation, is achieved primarily by the recirculation vortices and by the momentum flux ratio of the crossflow streams [63] [17].The conceptual micromix combustor design conceived at Cranfield University is shown as in Figure 2.

Figure 2 -Schematic front view of the Micromix Annular
Burner Segment

Micromix Design Parameters
The geometrical parameters for this micromix combustion design are essential to ensure the right penetration depth of hydrogen into the oncoming airstream.A single injector arrangement is shown as in Figure 3.The air stream flows through the micromix structure from left to right direction.The hydrogen jet penetrates perpendicular to the air stream shown in Figure 3.The turbulent generation and resulting eddy break down rapidly mixes the stoichiometric combustion zones formed locally of these micro-diffusion flames, hence reducing the local residence time at high-temperatures, and subsequent thermal NOx formation [16].
Recirculation vortices are required to anchor, stabilise and position the micro-flames over a range of engine operating conditions.The flames shape, size and position can then be adjusted by modifying the location and size of the recirculation zones downstream the micromix injector array.The presence of these recirculation areas are essential to ensure that the microflames are stable and are separated from each other in the vertical direction.It is critical to the design of the micromix system that these micro flames are separated from each other since the merging of adjacent flames could potentially result the formation of larger-scale flames.This would further increase fuel-air mixtures residence time in the hot recirculation regions and subsequently increase the NOx emissions, while potentially overheating the combustor liner [61].

Momentum Flux Ratio
The other main parameter that influences the micromix combustion system design is the momentum flux ratio between the fuel jet and the air crossflow.This momentum flux ratio is defined as [63]: Increasing the momentum flux ratio between the two streams increases the penetration depth of the hydrogen stream into the jet-in-cross-flow air stream thus enhancing the mixing characteristics [61][63].The air stream jet air mixes with the penetrating hydrogen stream, the inner recirculation zones are then divided by shear layers between the two streams.If the penetration depth is below a critical penetration height, the H2air mixture can be discharged freely, and the residence time is short.If, however, the penetration depth increases beyond this critical value, the hydrogen penetrates the shear layer between the two streams and hence enters into a hot inner recirculation zone [63], as shown in Figure 6.This further increases the residence time of the mixture, subsequently leading to an increase in NOx emissions [17].Hence the determination of this critical penetration depth (ycrit) is of vital importance to the design of the micromix system.The penetration depth and momentum flux ratio is dependent by equation (4):

Numerical Analysis
For the initial simulations of a single injector, a simplified numerical model of the geometry was solved for using ANSYS FLUENT™.A 3D numerical model of the micromix combustor section was created to analyse the reactive flow field using a RANS based solver along with a detailed H2/air reaction mechanism and a thermal NOx formation model.The resulting flow fields, temperature profiles, and OH species distribution across the computational domain were analysed.The influence of various geometrical design parameters on flame anchoring, its structure and hence the resulting emissions behaviour were also analysed.The aim of this preliminary numerical analysis was to understand the reacting flow characteristics of the jet-in-crossflow configuration and also to qualitatively identify the tendencies of varying injector designs.The analysis carried out as a part of this work would be used to inform and guide the design of the experimental rig to be developed as a part of the ENABLEH2 project at Cranfield [12].More details of the numerical model used for this study are stated below.

Computation Domain
The geometrical model of a single injector used this study is shown in Figure 11.It covers a longitudinal slice of a single injector configuration.It consists of an air inlet upstream of the burner geometry, the hydrogen inlet perpendicular to the air stream and the combustion volume.
Adiabatic wall with no-slip boundary conditions are initially used, instead of a symmetry boundary condition.This is done to isolate the effects of injector-injector interactions.The hydrogen and the air are injected through two discrete streams.The spatial discretisation of the computational domain consists of approximately 2.1 million hexahedral elements.The Φ used for this numerical study is assumed to be 0.43, an assumption based on the previous analysis carried out by Aachen [21].A detailed analysis on the effect of varying Φ on the hydrodynamic instability characteristics of this concept is required to derive the lean limit of hydrogen combustion at a given temperature and pressure condition and will be carried out in future work.The mass flow rate of the hydrogen is determined by the heat release required by each of the injectors for the micromix system.For the purposes of this study, the heat release per injector is assumed to be 3.74 kW and is kept constant.The combustor performance data and the boundary conditions imposed at the air and fuel inlet ports are summarised in Table 1.The fuel and air inlet boundary conditions used are representative of a typical ultra-high bypass ratio engine configuration for entry into service by the year 2050.The 3D steady RANS calculations have been performed using the k-ω SST turbulence scheme.The shear stress transport (SST) formulation allows for integration down to the wall, through the viscous sub-layer, resulting in a more accurate near wall treatment, which is vital for such designs involving small geometric features.This is particularly important when required to accurately capture a broad spectrum of eddies, ranging all the way from smaller scales of eddies in the hydrogen jet to the larger eddies in the recirculation zones [22].The primary challenges associated with turbulence chemistry modelling is to resolve the turbulent mixing scales along with combustion chemistry.This may involve solving hundreds of molecular species, in a solution time compatible with the engineering design cycle time [23].Hence, the combustion reaction kinetics for the present work are modelled using the FGM approach for a partially premixed diffusion flamelets [24][25].FGM is an industry standard approach used to reduce the computational expenses associated with directly solving the reaction kinetics.The FGM modelling approach is based on the fact that in many combustion processes, the chemical, or reaction time scales occur much faster than the actual flow time scales.These fast chemical processes can therefore be decoupled.This reduces the stiffness of the numerical model and therefore reduces the computational costs [24].The turbulent flame is assumed to be an ensemble of laminar flames, the internal structure of which is not significantly influenced by the flow turbulence.These flamelets are then embedded into the turbulent flame brush using statistical averaging.The species and associated thermodynamics are parameterised by just a few variables such as reaction progress, scalar-dissipation and mixture fraction.A set of transport equations are then solved for these variables.FGM is fundamentally different from other laminar flamelet models primarily because of its capability to model non-equilibrium effects.Since these other laminar flamelets models are parameterised by strain, the thermochemistry always tends to equilibrium as the strain rate decays towards the exit of the combustor.In contrast, the FGM model is parameterised by the reaction progress variable; therefore, the flame can be fully quenched.No assumptions of thin and intact flamelets are made by the FGM.This is particularly important when modelling ignition and extinction phenomenon.The flamelet library is created using the ANSYS FLUENT 19.1 diffusion laminar flamelet generator [25].Many works exist in the literature stating the limitations, the applicability and the predictive capabilities of using FGM to model the combustion reaction kinetics [25-27] [60].
The default number of flamelets that can be generated has been capped to 100 in ANSYS FLUENT.Due, to the high pressure and temperature conditions for the present numerical simulations, the original set of hydrogen flamelets were observed to be extinguished at a strain rate of 7851/s, and a scalar dissipation increment of 79/s had to be used to respect the default number of flamelets.However, such high scalar dissipation step may lead to poor interpolation within the PDF lookup table and therefore affect the predictive capabilities of the combustion kinetics.To reduce the scalar dissipation step size to a more reasonable size, a scheme has been developed and implemented in a FLUENT to increase the default number of flamelets to 1600.Scheme is a LISP dialect used to develop the Fluent GUI, and is often used to interact with the FLUENT journal files via parameters.By the implementation of scheme the scalar dissipation steps can be reduced to 5/s, allowing for a much finer discretisation of the flamelet space.
The kinetics of the thermal NOx formation rate is much slower than the compared to the energy-releasing reactions, and so most of the thermal NOx is formed only after the complete combustion.Theoretically, the FGM modelling approach can be used to simulate the slow chemistry of NOx formation but the degree of uncertainty associated with it is very high [24].For the purposes of this analysis, the thermal NOx predictions are estimated, in a post-processing exercise, using the extended Zeldovich mechanism on the frozen reactive flow field predicted by the FGM model.

Combustion Reaction Mechanism Comparison
There are several mechanisms available in the literature for hydrogen combustion.Each of these mechanisms has been developed under varying experimental types and conditions.Carsten [28] has carried out a comprehensive comparison of the published hydrogen combustion mechanisms.For the purposes of this study three mechanisms have been down-selected based on previous work carried out at Cranfield [29].The previous work at Cranfield was limited to the analysis under atmospheric pressure conditions only.It is known that although hydrogen has a relatively simple reaction mechanism, the mechanism is also highly dependent on the pressure conditions.Therefore, the down selection of hydrogen-air mechanisms under high-pressure conditions was carried out.The manifold represents the chemistry inside a three-dimensional CFD simulation.The turbulence-chemistry interaction is pre-tabulated for the entire manifold and therefore requires no added computational efforts.This enables the use of detailed combustion chemistry mechanism with no added computational costs.
Although various mechanisms are available in the literature [30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46][47][48], there are relatively very few mechanisms that have been derived under high-pressure experimental conditions.The mechanisms analysed are shown in Table 2. GRI-Mech 3.0 [43] is used in various industrial applications for simulations and it covers hydrogen and many hydrocarbon combustion reactions.DRM 1.2 [49] was developed as a truncated version of the GRI-Mech.This was done to develop a smaller subset of reactions capable of closely producing the main combustion characteristics as predicted by the complete mechanism.CV Naik [50] mechanism, on the other hand, is developed to simulate the combustion of pure H2 and O2 mixture, over wider range of pressure ranging from 0.05 atm to 87 atm, Φ from 0.2 to 6 and temperatures ranging from 298K to 2700K.

Laminar Burning Velocity
Laminar burning velocity is considered to be among the most important combustion parameters used in assessing many phenomena like ignition, quenching of the flame, and its flashback stabilisation [51].LBV vs Φ of the mechanisms is shown as in Figure 7. Laminar burning velocity of an H2 and air mixture is generally much higher than for conventional fuels, an outcome associated with the high reactivity of H2/air.Generally, the same laminar burning velocity is predicted by all the three mechanisms, over a wide range of Φ as shown in Figure 7.However, at very high values of Φ (>5) and high-pressure conditions, the GRI-Mech 3.0 tends to overpredict the laminar burning velocity by about 40-50 % compared to the other two mechanisms as shown in Table 3.Such variations are observed at very lean conditions as well.Similar discrepancies have been reported under lean high-pressure conditions by Goswami [59] where the researcher concluded the unsuitability of using the GRI-Mech3.0mechanism for pure hydrogen combustion under such high pressure and lean conditions.

OH Radical Distribution
The concentration of OH species is of prime importance in a hydrogen combustion system and is often employed as a flame marker.Among the various reaction steps one of the most important chain propagation reaction steps of hydrogen reaction, and in NOx chemistry, is the reaction between the molecular oxygen and hydrogen radical to form hydroxyl and oxygen radicals (H + O2 = OH + O) and the reaction of an oxygen radical with molecular hydrogen to form hydrogen and hydroxyl radical (O + H2 = OH + H).The reaction rate coefficients of these reaction steps are highly dependent on the conditions under which the experimental analysis was carried out [53].Large number of experimental analysis have been carried out to determine the rate constants of this specific reaction step; literature [53] suggests that significant disagreement still prevails in their determination.The OH species distribution along the computational domain is shown as in Figure 8.The variations of OH radicals predicted by each of these mechanisms can perhaps be attributed to the experimental conditions under which the reaction rate constants were determined.

Temperature Distribution
The mass-weighted temperature distribution along the midsection of the computational domain is shown in Figure 9.All the three mechanisms tend to follow the same temperature trend across the domain.However, upon closer analysis, the peak temperatures predicted by these mechanisms show important variations with peak temperatures of 2648 K, 2603 K and 2545 K predicted by DRM 1.2, Naik, and GRI-Mech3.0respectively.This may be attributed to variations in the formation of the OH radical predicted in the reacting domain, since the same trends observed for the peak temperature distributions are also observed for the OH radical distributions.The formation of OH radical in the hydrogen reacting system is an exothermic reaction process.The discrepancies in the prediction of hydroxyl radicals can, therefore, lead to variations in the peak temperature distribution and consequently variations in NOx chemistry.

NOx Distribution
The presence of high-temperature peaks will further influence the chemistry of the thermal NOx formation.In fact, the thermal NOx formation doubles for every 90K temperature increase beyond 2200 K [54], hence the trends shown in Figure 10.The NOx values shown in Figure 10 are normalized against the peak area averaged NOx value predicted by GRI-Mech3.0mechanism, this is due to the uncertainty in the absolute value of the NOx emissions predicted by the current numerical model as stated previously.
Both the GRI-Mech3.0and DRM 1.2 mechanism have been highly optimised for natural gas combustion and perhaps because of this, it fails to predict the hydrogen reaction accurately.In fact, various researchers have questioned the applicability of the GRI-Mech 3.0 for hydrogen combustion [28] [52].
Based on this analysis, the Naik mechanism was downselected to be used for further analysis.The reasons for the uncertainties associated with these mechanisms are highly debatable and hence require further experimental validation which would be carried out as a part of future work.

Micromix Combustion Reactive Flow Field
The driving design factor of a jet in crossflow is to obtain a homogenous mixture rapidly of the injected fuel and the free stream air.The degree and the rate of the mixing process are important as it affects the combustion process.The burning efficiency and exhaust compositions are directly depend upon the reaction kinetics and mass transfer.The chemical kinetics are difficult to control but, the mixing process is more readily controlled by many numbers of parameters.The flame obtained by the present micromix combustor concept is anchored, near the edge of the hydrogen injector segment, by two recirculation zones and is characterised by two distinct reaction zones, as shown in Figure 4.The outer zone is formed by the recirculation of hot combustion products downstream the H2 penetration site and the inner zone is formed by the recirculation of the air downstream the air feed panel as shown in Figure 11.Shear layers is established between the two recirculation zones because of the shift in the axial position of the H2 stream and the reaction zone is established along the inter-recirculation shear layer.The orientation and the structure of these micro-flames are in fact influenced by the shear layer.This shear layer is in turn defined by the size and the position of the recirculation vortices [15].Here the remaining unburnt H2 from the first flame region starts to react.This kind of flame structure isn't typical of a micromix combustor.The micromix principle aims to combust all the injected fuel in the thin inter-vortex shear layer thus avoiding high fuel concentration and, high-temperature peak and consequently reduced NOx formations [61] [62].

Parametric Analysis
The various geometrical parameters that influence the design of the micromix injector are shown in Figure 3.Each of these design variables influence the mixing characteristics between the fuel and oxidiser streams therefore the NOx emissions.One of the primary factors that influence the design of the micromix system is the momentum flux ratio between the fuel and oxidiser stream.As a part of this study, the influence of varying the air feed panel hydraulic diameter and the hydrogen inlet diameter and their subsequent impact on the momentum flux ratio and NOx emission characteristics are analyzed.A reference geometry is scaled based on the previous work carried out by Aachen [17], such that the momentum flux ratio was kept constant but with an increased heat release for the current analysis.The reference geometry design parameters are shown in Table 3.With the reference geometry shown in Table 3 the following parametric studies have been carried out: 1. Variation in the hydrogen inlet diameter: The hydrogen inlet diameter is varied ranging from 0.183 mm to 0.5 mm.The geometrical model is parametrised using ANSYS scripting, such that any variations in the hydrogen inlet diameter would not influence the mixing distance and the offset distance between the hydrogen and the air stream.This is done to isolate the influence of a single parameter variation i.e. the hydrogen inlet diameter on the mixing and hence the emission characteristics.2. Variations in the air feed panel hydraulic diameter: The air feed hydraulic diameter has been varied ranging from 1.83 mm to 3.34 mm.The aspect ratio of 1.5 for the air feed panel is kept constant for all the analysis.The mixing and the offset distance are both kept constant.The height of the computational domain is scaled proportionally to the variation in the air feed hydraulic diameter, this is done to reduce the impact of hydrogen feed height on outer recirculation zone.The hydraulic diameter of the air feed panel and the diameter of hydrogen inlet is varied such that the momentum flux ratio ranges from 0.5 to 20.

RESULTS AND DISCUSSION
The influence of varying hydrogen injector diameter on the NOx distribution is shown in Figure 14.All the measurements for the NOx emission were taken at plane x=70 mm downstream the injector segment.The location of the measurement plane was selected such that it lies downstream the established mixing region.The plot is normalised using the mass-weighted average NOx predicted at x =70 mm from the reference geometry.From Figure 14 with a decrease in hydrogen inlet diameter, and the associated increase in the momentum flux ratio and mixing characteristics, there is a net improvement in predicted NOx formation.However, beyond a certain decrease in the hydrogen inlet diameter, the NOx emissions begin to increase.This can be attributed to the penetration of hydrogen into the outer recirculation zone as shown in Figure 6.
Similarly, the influence of varying the air feed panel hydraulic diameter on the NOx characteristics is shown in Figure 15.From Figure 15 with an increase in air feed hydraulic diameter and the associated increase in momentum flux ratio and hence the mixing characteristics, there is a net improvement in the predicted NOx formation.This can be related to the jet-incrossflow momentum flux ratio and to the recirculation zone that surrounds and stabilises the H2 flames.An increase in the cross flow momentum flux ratio results in a greater penetration depth as shown in Equation 4. Upon further increasing the momentum flux ration beyond a critical value, the H2 stream penetrates the shear layer between the cross-flow streams and consequently increases the NOx emissions.

8
Copyright © 2019 ASME The temperature and NOx contour plots along the mid-section of the computational domain for the best and the worst case scenarios are shown in Figure 16 and Figure 17.The dimensions of the "preferred injector" are shown in Table 4.The calculated NOx shown in Figure 17, reflects the flame structure shown in Figure 16.The low momentum flux ratio results in reduced penetration of the hydrogen into the oxidiser stream.This results in a greater amount of heat release occurring at the lower side of the flame as shown in Figure 16.However, increasing the momentum flux ratio between the jet streams results in more penetration of the fuel into the oxidiser stream.This results in the heat release occurring in the shear layers between the outer and inner recirculation zones.This is shown as in Figure 16.The flame zone length is also reduced, which further results in decreased peak temperature regions and hence reduced NOx formation regions as shown in Figure 17.
From Figure 16 it is observed that the anchoring position of the hydrogen flame has been impacted because of the variation in the air feed hydraulic diameter and the hydrogen inlet diameter.Upon increasing the momentum flux ratio, the distance between the recirculation zone and the injected H2 jet is reduced.If the distance becomes too marginal, the anchoring position of the hydrogen flame is altered.The flame anchors onto the shear layer of the air feed panel recirculation zone instead of the mixing segment of the burner [63].The can therefore possibly have implications on the stability of the flame [59] [61].It must also be emphasized that the peak NOx values predicted for all the injector cases are higher than those investigated in the literature [8] [17].The previous analyses carried out on micromix combustion systems have been limited to its application in low power systems.However, the current investigation is carried out for the implementation of the micromix combustion technology for high thrust class engines hence an increase in heat release per injector.Due to this increase in heat release per injector the mass flow rates for both the fuel and the oxidiser have increased which further leads to thicker and longer micromix flames and hence greater NOx emissions.

Figure 3 -Figure 5 -
Figure 3-Design Variable for Single Injector Study