Hydrogen-assisted catalytic ignition characteristics of propane–air with a chemical kinetic model in a Pt/γ-Al₂O₃ micro-combustor in different feeding modes

Abstract

From ambient cold-start conditions, the hydrogen self-ignition and hydrogen-assisted ignition of propane–air mixtures with a chemical kinetic model in different feeding modes were investigated numerically in Pt/γ-Al₂O₃ catalytic micro-combustors. For the steady and transient state, the micro-combustion and self-ignition characteristics of lean hydrogen–air mixtures were presented, and the hydrogen-assisted combustion of propane–air mixtures was investigated numerically in the co-feed mode and the sequential feed mode. The computational results indicate the large thermal inertia of the micro-combustor solid structure leads to slow temperature dynamics, and the transient response is dominated by the thermal inertia. In general, the concentration of hydrogen required for propane ignition increased with increasing wall thermal conductivity, decreasing inlet velocity, and decreasing inlet equivalence ratio of propane–air mixtures. In the co-feed mode, the combustion characteristics of hydrogen-assisted propane qualitatively resemble the selectively preheating initial portion of the combustion chamber wall. In the sequential feed mode, the time taken to reach the steady state, the hydrogen cut-off time, the propane ignition time and the cumulative propane emissions increased with increasing wall thermal conductivity; the ignition characteristics are similar to partially preheating the initial segment of the micro-combustor for low and moderate wall thermal conductivity values (0.5 and 20 W m⁻¹ K⁻¹); however, the ignition characteristics are close to completely heating the micro-combustor wall for high wall thermal conductivity values (200 W m⁻¹ K⁻¹). The minimum cumulative amount of hydrogen usage and minimization of startup time are discussed.

---

1. Introduction

Micro-combustors are increasingly investigated for the non-catalytic and catalytic portable production of heat and/or energy.¹ The produced energy of micro-combustion can be utilized by various methods. Examples include heat supply to micro-reactors carrying out endothermic reactions, such as steam reforming or ammonia decomposition to produce hydrogen for portable fuel cells,² and heat generation for remote use,³ such as for space flights and soldiers, thermo-electrics to produce electricity.⁴ Because hydrocarbons and hydrogen possess a significantly greater energy density than the traditional lithium-ion and metal acid batteries⁵, hydrocarbon-based micro-combustors are enticing prospective energy sources for portable power applications, such as portable electronics, laptops, cell phones, and personal heaters.⁶

As the characteristic length scales of combustors decrease, the surface-to-volume ratio increases, leading to enhanced mass and heat transfer rates between the surface and fluid, making micro-combustors susceptible to radical and thermal quenching on the walls.^7,8 The critical dimension is comparable to the quench distance of micro-combustors, which makes homogeneous combustion unstable at micro-scales. In order to improve micro-combustion stability and thermal efficiency, various efforts, such as catalytic combustion and pre-mixed combustion,^9,10 heat-recirculating^11,12 and “swiss roll”^13,14 combustors, have been paid on the thermal management optimization. Meso-scale combustion has emerged including the heat-recirculating and “swiss roll” combustors with the gap sizes of the order of 3.0 mm, for heterogeneous and homogeneous combustion of propane–air mixtures.¹⁵

A significant aspect of catalytic micro-combustion of hydrogen over the noble metal catalysts (such as Pt, Pd, Rh, etc.) is that the reaction rates are extremely fast and the reaction has low activation energy E_a. The above-mentioned property can be employed for igniting combustion of hydrocarbons, which referred to as hydrogen-assisted ignition.¹⁶ Norton and Vlachos¹⁷ explored the hydrocarbons ignition by hydrogen-assisted catalytic combustion in confined ceramic micro-channels with Pt catalyst. They found that the minimum hydrogen composition for self-ignition of propane/air mixture compositions is relatively constant: minimum hydrogen concentration of ∼3.6% (mole fraction), irrespective of propane composition. Moreover, they described the steady and transient state behavior, discussed the minimization of hydrogen usage and startup time, and compared different ignition procedures (co-feed and swapping mode). Their results indicated that the reactivity of hydrogen is inhibited by propane due to kinetics, higher H₂ compositions lead to a relatively fast startup and minimum H₂ utilization, and co-feeding hydrogen with propane is a good startup strategy. Yan et al.¹⁸ demonstrated the self-ignition behavior of hydrogen, and numerically investigated hydrogen-assisted ignition of methane–air mixtures in catalytic micro-combustors with a channel gap of 2.2 mm. The results showed that approximately 3% hydrogen by volume is sufficient to ignite propane–air mixtures at atmospheric temperature. Zhong et al.¹⁹ experimentally investigated hydrogen-assisted and external heating catalytic ignition characteristics of n-butane in a Pt-coated monolith catalytic micro-reactor, and discussed two startup methods and thermal insulation. They found that the co-feed method, large hydrogen mole fraction and thermal insulation are beneficial to catalytic hydrogen-assisted ignition. Furthermore, they also experimentally and numerically investigated the catalytic ignition processes and kinetics of hydrogen-assisted n-butane on Pt surface to reveal the catalytic ignition mechanism.^19–21 The results showed that the effect of hydrogen on the catalytic ignition process and temperature of n-butane depends on the concentration of hydrogen added. During hydrogen-assisted catalytic ignition of n-butane–air mixtures, hydrogen not only has a thermal effect, but a chemical effect as well.¹⁹ The thermal effect is exhibited when small fractions of hydrogen are added. However, it is changed into chemical effect as the concentration of hydrogen is increased.²⁰ The concentration ranges of hydrogen with different effects were also predicted. The critical concentration value of hydrogen required for n-butane catalytic ignition is 0.025 on a molar basis.²¹ Deutschmann et al.²² experimentally and numerically investigated hydrogen-assisted catalytic combustion of methane–air mixtures on platinum coated honeycomb channels. They found that the light-off is mainly determined by the catalyst temperature because of the heat release due to H₂ catalytic combustion; and increasing H₂ addition ensures the light-off, decreasing H₂ addition requires an increasing CH₄ feed for light-off. Furthermore, the results showed that the main effect of hydrogen is thermal: hydrogen ignited firstly on the surface of catalysts, then the temperature of mixtures increased from ambient temperature to ignition point. In their work, the chemical effect of hydrogen was not discussed.

Most current prototypes of micro-combustors depend on exterior heating to generate energy for ignition, and the additional equipment necessary to power the exterior heaters can negate mass advantages of micro-combustors. In this work, the hydrogen self-ignition and hydrogen-assisted ignition of propane–air mixtures in different feeding modes are investigated numerically in catalytic micro-combustors. The effect of catalytic chemical kinetics on self-ignition of micro-combustors is delineated. Moreover, for the steady and transient state, a reduced-order reaction model is employed to examine the ignition characteristics of hydrogen at micro-scales. Finally, the hydrogen-assisted combustion of propane–air mixtures is investigated numerically in different operation modes of fuel feed.

2. Numerical models and simulation approach

2.1. Model geometry and mesh

A schematic view of the catalytic micro-combustor modeled in this work is shown in Fig. 1. The kinetics module plug-in for FLUENT was employed to simulate the flow of hydrogen–propane–air mixtures in the plane channel of height H = 0.2 mm, length L = 20.0 mm, and solid wall thickness δ = 0.2 mm. The wall material is refractory ceramics SiC, which thermal conductivity k_s, the emissivity ε, density ρ and specific heat capacity c are 20 W m⁻¹ K⁻¹, 0.8, 3.2 × 10³ kg m⁻³ and 800 J kg⁻¹ K⁻¹, respectively. Inner horizontal surfaces of micro-channel contained Pt/γ-Al₂O₃ catalyst washcoat. The properties of Pt/γ-Al₂O₃ supported noble metal catalysts are shown in Table 1. For all scenarios analyzed, the uniform grid size of 0.005 mm was used to mesh the computational domain.

Fig. 1 Schematic diagram of the micro-channel geometry.

Table 1 The properties of Pt/γ-Al₂O₃ catalyst washcoat

Property	Value
Catalyst surface site density Γ (mol cm^−2)	2.7 × 10^−9
Average pore diameter dpore (m)	2.08 × 10^−8
Catalyst porosity εcat	0.4
Catalyst tortuosity τcat	8.0

---

In the present work, in order to couple the hetero-/homogeneous chemical kinetics, fluid dynamics, and heat transfer, FLUENT-KINETICS was employed to simulate the chemical kinetics. The residuals were below 10⁻⁵ for all simulations.

2.2. Chemical kinetics for hydrogen combustion

The chemical kinetics for the micro-combustion of hydrogen–air is adopted from Vlachos,²³ who developed the model via the hierarchical model reduction of 13-step surface catalytic reaction mechanism. The resulting reaction rate of micro-combustion is given by the following reduced-order reaction rate kinetics:where the effectiveness factor η is 1, the pre-exponential factor A₀ is 1280 cm K^−0.5 s⁻¹, the temperature exponent β is 0.5, and the activation energy is 0 kJ mol⁻¹ because of hydrogen adsorption is the non-activated process.²³

Vlachos²⁴ reported that homogeneous combustion of near-stoichiometric hydrogen–air mixtures occurs in the larger gap size of 1.0 mm, but not for the smaller gap size of 0.25 mm. Even in the 1.0 mm micro-combustor, hydrogen light-off in gas phase was observed beyond the equivalence ratio φ_H2 of 0.33. Moreover, Vlachos²⁴ also reported that the above-mentioned kinetic model is valid for the concentration of hydrogen in the range of the equivalence ratio 0.008 < φ_H2 < 0.424. Mantzaras²⁵ observed hydrogen light-off in gas phase for the equivalence ratio φ_H2 of 0.2 in 1.2 mm diameter micro-channels. Therefore, in this study, homogeneous combustion of hydrogen–air mixtures is neglected and Numerical simulations are restricted for hydrogen concentrations below the equivalence ratio φ_H2 of 0.2.

2.3. Chemical kinetics for propane combustion

In the present work, Pt/γ-Al₂O₃ catalysts were adopted for catalytic micro-combustion of propane–air mixtures. The reason is that the propane conversion of the noble metal catalysts decreases in the order Pt/γ-Al₂O₃ > Pd/γ-Al₂O₃ > Rh/γ-Al₂O₃ at the stoichiometric propane–oxygen ratio.²⁶ The reduced-order reaction rate kinetics for catalytic micro-combustion of propane–air mixtures are as follows:where r_cat,C3H8 is the surface catalytic reaction rate of propane. C_s,i is concentration of adsorbed species i. k_i is adsorption or desorption rate constant of species i, and determined as follows:where s₀ is the sticking coefficient. Γ is Pt/γ-Al₂O₃ catalyst surface site density. M_i is the molecular weight of species i. T_ref is the temperature of reference conditions. The values of kinetic parameters for the catalytic combustion of lean propane are shown in Table 2.

Table 2 Kinetic parameters for the catalytic combustion of lean propane over Pt/γ-Al₂O₃

	A0 (s^−1) or s0	β	Ea (kcal mol^−1)
a , surface area factor η = 1.7, the temperature of reference conditions Tref = 300 K.
C3H8 adsorption	0.06	0.154	4
O2 adsorption	0.0542	0.766	0
O2 desorption	8.41 × 10^12	−0.796	^a

---

The activation energy of oxygen desorption depends on the coverage θ_O˙ of oxygen radical, which is calculated as follows:

2.4. Boundary conditions

The incoming flow of hydrogen–air or hydrogen–propane–air mixtures was fully premixed and have uniform inlet temperature T_in of 300 K. The thermal boundary condition on the wall is the heat loss to the ambient air. The heat losses to the surroundings are calculated using the following equation

q = h(T_w,o − T_∞) + εδ(T_w,o⁴ − T_∞⁴) (6)

where the exterior convective heat transfer coefficient h is 20 W m⁻² K⁻¹ in this study. T_w,o is the temperature at the exterior wall surface. The ambient temperature T_∞ is 300 K. The solid wall emissivity ε is taken to be 0.8 and δ is the Stefan–Boltzmann constant in eqn (6).

3. Micro-scale ignition characteristics of hydrogen

3.1. Steady state behavior

The reaction rate of H₂ micro-combustion on Pt/γ-Al₂O₃ catalyst is very fast. Despite the high diffusivity of hydrogen, the catalytic combustion of hydrogen–air mixtures at micro-scales is diffusion limited.^27–32 In the present work, the incoming hydrogen–air mixtures was fully premixed with the equivalence ratio φ of 0.2. The effect of inlet velocity V_in on the temperature along central axis and interior wall as well as the H₂ conversion for adiabatic and non-adiabatic micro-combustors at steady state are shown in Fig. 2–4. The micro-combustion occurs close to the inlet section and only a fraction of micro-combustor length is required to attain complete H₂ conversion, even for higher velocity case. The maximum temperature T_max is attained at the beginning of the interior wall, by reason of the surface catalytic reaction. As the inlet velocity V_in is increased or heat transfer coefficient h is decreased, the maximum temperature T_max on the interior wall increases. In the case of non-adiabatic micro-combustors, the upstream interior wall gets heated up because of the surface catalytic reaction, whereas the net heat loss to the ambience occurs in the downstream section. Therefore, for the moderate to high heat losses case, the direction of heat transfer in micro-combustors is from the solid wall to the gas phase in the upstream section as a result of the surface catalytic reaction, whereas the direction reverses in the downstream section by reason of the heat losses. In adiabatic micro-combustors or at higher inlet velocity (low residence time), the above-mentioned reversal is not observed. The wall thermal conductivity also affects the overall temperature distribution in micro-combustors. For lower wall thermal conductivities, the significant temperature gradients exist as observed, whereas the interior wall temperature profiles are nearly flat for the highly conducting walls in Fig. 5 and 6.

Fig. 2 Temperature along central axis for adiabatic and non-adiabatic micro-combustors at steady state.

Fig. 3 Temperature along interior wall for adiabatic and non-adiabatic micro-combustors at steady state.

Fig. 4 H₂ conversion for adiabatic and non-adiabatic micro-combustors at steady state.

Fig. 5 Transient evolution of the central axial temperature profiles in the gas phase for H₂ catalytic ignition for different wall thermal conductivities. φ (H₂) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 6 Transient evolution of the interior wall temperature profiles on the wall for H₂ catalytic ignition for different wall thermal conductivities. φ (H₂) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

3.2. Transient behavior

In the present work, the primary purpose is to analyze the transient response of the micro-combustor. The micro-combustor initially contains only air and starting at ambient conditions of 300 K. The premixed hydrogen–air mixtures is fed starting at the time t = 0 and the desired inlet velocity V_in (4 m s⁻¹) and equivalence ratio φ (0.2). In order to ensure the reaction system reaches steady state, numerical simulations are run for enough time. Norton and Vlachos³³ numerically investigated the effects of micro-combustor wall conductivity, external heat losses, operating conditions, and micro-combustor dimensions on combustion characteristics and the steady-state, self-sustained flame stability of propane–air mixtures. They found that the wall thermal conductivity is vital in determining the flame stability of the micro-combustor, as the walls are responsible for the majority of the upstream heat transfer as well as the external heat losses. Therefore, the effect of the wall thermal conductivity on the ignition characteristics of propane–air are investigated in this work.

For low (0.5 W m⁻¹ K⁻¹), moderate (20 W m⁻¹ K⁻¹) and high (200 W m⁻¹ K⁻¹) thermal conductivities k_s of wall material, the temporal responses of the central axial and the interior wall temperature profiles from ambient cold-start conditions are shown in Fig. 5 and 6, respectively. As observed in Fig. 5 and 6, the thermal conductivities k_s of wall material have significant effect on the temperature distribution in the micro-combustors; in the bulk gas (Fig. 5) and the wall (Fig. 6), the large temperature gradients exist in low thermal conductivity walls whereas the temperature profiles are nearly flat for the highly conducting walls. However, the H₂ conversion is not affected by the thermal conductivities of wall material. In all cases, front-end ignition in micro-combustors is observed and H₂ conversion attain the steady state profile in Fig. 4 in less than 1.0 s. Due to the H₂ conversion profile does not change in micro-combustors for most of the simulation time after about 1.0 s, the above-mentioned behavior is often referred to as pseudo-steady-state. Compared with the diffusion and reaction processes, the large thermal inertia (which is defined as the square root of the product of the material's bulk thermal conductivity and volumetric heat capacity, where the latter is the product of density and specific heat capacity) of the micro-combustor solid structure leads to slow temperature dynamics, and transient response is dominated by the thermal inertia.^34–36

From the cold-start conditions, the variations of the outlet temperature in gas phase for different wall thermal conductivities are shown in Fig. 7. As observed, the time when the micro-combustor reaches steady state is denoted by symbols in Fig. 7. The symbols time is defined at which the maximum deviation of the outlet temperature is less than 0.5 K from the steady state values. As wall thermal conductivity is increased, the steady state time increases. For low wall thermal conductivity, a high temperature gradient exists on the wall; a high temperature gradient on the wall will make the homogeneous combustion shift upstream and the micro-combustor will have a higher peak temperature.³⁷ On the contrary, for high wall thermal conductivity, a low temperature gradient on the wall will make the homogeneous combustion shift downstream and the micro-combustor will have a higher outlet temperature (as observed in Fig. 7). For different inlet velocities and wall thermal conductivities, the time taken to reach steady state are shown in Fig. 8. As the inlet velocity is increased, the steady state time decreases due to the maximum wall temperature increase and the net energy supplied. As observed in Fig. 7 and 8, for micro-combustor with lower wall thermal conductivity, the steady state time is shorter. The reason is that the heat localization in poorly conducting walls leads to fast ignition. At low inlet velocity, the effect of wall thermal conductivity on the steady state time is especially significant; as the inlet velocity is increased, the effect diminishes as a result of the increased convective heat transfer.

Fig. 7 Outlet temperature in gas phase versus time for different wall thermal conductivities. The symbols represent the time when the micro-combustor reaches steady state.

Fig. 8 The entire time required to reach steady state from initial conditions for different inlet velocities and wall thermal conductivities.

In general, the catalytic combustion in micro-combustors reaches pseudo-steady state instantaneously; the combustion region is a short zone near the inlet section; and the higher inlet velocities lead to faster transient response.

4. Micro-scale ignition characteristics of hydrogen-assisted propane

Hydrogen combustion can be employed to ignite propane–air mixtures, due to the wall temperatures reached in hydrogen–air combustion are higher than the ignition temperature of propane. Which is referred to as hydrogen-assisted ignition of propane.^{3,4,16–18,27} Norton and Vlachos¹⁷ compared ignition characteristics of hydrogen-assisted propane at different equivalence ratios of propane–air (0.1 and 0.04). Furthermore, they compared different ignition procedures:

(a) Co-feed mode: the co-feed method entails simultaneously turning on the flow of hydrogen, propane, and air until the propane ignites, and then turning off the hydrogen flow and increasing the flow rates of propane and air so that the total flow rate remains constant;

(b) Swapping mode: one preheats the micro-combustor using hydrogen–air–nitrogen combustion, then switches the propane on and the nitrogen off when the ignition temperature of propane has been reached, and finally turns off the hydrogen flow while increasing the propane and air flow rates to maintain constant of the total flow rate.

They found that higher hydrogen compositions (0.1) lead to a relatively fast startup of the micro-combustor and minimum hydrogen utilization while being safe, although small fractions (0.04) of hydrogen are adequate to cause the self-ignition of propane. Moreover, their results indicated that the co-feed mode (simultaneous starting of flows) is simple and has the additional benefit of heat release from both fuels, so it is a good startup strategy. The experimental results from the effect of hydrogen-assisted on catalytic ignition characteristics in co-feed and step-feed modes by Zhong et al.¹⁹ also validated the above-mentioned conclusion, although the fuel is n-butane, and step-feed mode (hydrogen–air mixtures are supplied first, and n-butane is supplied when the characteristic temperature reaches sufficiently high temperature) is different from swapping mode.

In the present work, transient response of different feed modes is analyzed: (a) co-feed mode (premixed hydrogen–propane–air mixtures); (b) sequential feed mode (switched from hydrogen–air to propane–air mixtures).

In the sequential feed mode, hydrogen–air mixtures at the desired equivalence ratio are fed into the micro-combustor firstly. Once the sufficiently high temperature (ignition temperature of propane) in micro-combustor is reached, the hydrogen feed is cut off and the feed of propane–air mixtures is started. Simultaneously, keep total flow rate constant.

Note that the sequential feed mode in this work is different from the above-mentioned swapping mode: (a) initial feed: hydrogen–air vs. hydrogen–air–nitrogen; (b) switching method: switch from hydrogen to propane vs. switch from nitrogen to propane; (c) switching sequence: turn off hydrogen then turn on propane vs. turn on propane then turn off hydrogen. The purpose of nitrogen added by Norton and Vlachos¹⁷ was to investigate the kinetics effect of propane on hydrogen catalytic combustion. In the present work, the sequential feed mode is simple and easy to operate comparing with the swapping mode.

Moreover, the sequential feed mode is also different from the above-mentioned step-feed mode: (a) turn off hydrogen when the ignition temperature of propane has been reached vs. hydrogen is always turned on; (b) keep total flow rate constant vs. the flow rate of hydrogen and air is maintained constant while the total flow rate is increased because of adding the second fuel (n-butane).

In the present work, the thermal effect of hydrogen on hydrogen-assisted catalytic ignition is the main influence factor, because the concentration of hydrogen added is low (Zhong et al.^19–21). Therefore, the chemical effect is not consider in this work. Due to absence of detailed reaction mechanism of hydrogen-assisted catalytic ignition of propane–air mixtures, the variation of free radicals and the effect of hydrogen on the ignition temperature of propane are also not investigated in this work.

4.1. Premixed hydrogen–propane–air mixtures (co-feed mode)

4.1.1 H₂ requirement for C₃H₈ ignition. In micro-combustors, the bifurcation behavior of C₃H₈ catalytic ignition in hydrogen-assisted co-feed mode are shown in Fig. 9. The inlet velocity is set to 10 m s⁻¹. The equivalence ratio φ of premixed propane–air mixtures is kept fixed (0.6 and 0.7). As the hydrogen mole fraction is increased starting from a zero value, the flow rates of propane and air are decreased, while the hydrogen flow rate increases to maintain the total volumetric flow rate constant. As H₂ mole fraction is increased, the maximum temperature in micro-combustors increases gradually because of H₂ combustion in the feed. Then, after a particular mole fraction of hydrogen, the heat released on H₂ combustion is sufficient to ignite C₃H₈ combustion and the turning point bifurcation is observed in Fig. 9. The above-mentioned bifurcation behavior is consistent with the previous work.¹⁷ For comparison, the ignition temperatures in micro-combustors in inlet feed preheating mode are also shown in Fig. 9. For the hydrogen-assisted ignition case, the maximum wall temperatures at the ignition bifurcation are slightly higher than the inlet preheating case. As observed in Fig. 9, the higher equivalence ratio of propane–air mixtures leads to earlier ignition.

Fig. 9 The ignition turning point bifurcation in micro-combustors for catalytic propane–air ignition in hydrogen-assisted co-feed mode (black and red lines, bottom abscissa) and inlet preheating mode (green and blue lines, top abscissa).

For different inlet equivalence ratios of propane–air mixtures, the effect of inlet equivalence ratio on the H₂ requirement for self-ignition is shown in Fig. 10. The inlet velocity is set to 10 m s⁻¹. For comparison, the C₃H₈ ignition temperatures without H₂ addition in inlet feed preheating mode are also shown in Fig. 10. As observed, as the inlet equivalence ratio of propane–air mixtures is increased, the H₂ requirement in the feed reduces. This behavior is consistent with the conclusion from Deutschmann et al.²² (the mole fraction of hydrogen needed for ignition decreases when hydrocarbon concentration is increased). For the hydrogen-assisted ignition case, the above-mentioned trend is qualitatively similar to the inlet preheating case.

Fig. 10 The concentration of H₂ required in micro-combustors for ignition in hydrogen-assisted co-feed mode (black line, left ordinate) and the ignition temperature (red line, right ordinate) for different equivalence ratios of propane–air mixtures.

As observed in Fig. 11, inlet velocity and wall thermal conductivity both have strong effect on the concentration of H₂ required for C₃H₈ ignition. As the inlet velocity is increased, the wall temperatures increase (see Fig. 3). Therefore, as the velocity is increased, the concentration of H₂ required for C₃H₈ ignition decreases. Likewise, as the wall thermal conductivity is decreased, the concentration of H₂ required for C₃H₈ ignition decreases. The reason is that the hot spots created near the inlet section of micro-combustors for lower thermal conductivity material as observed in Fig. 2 and ref. 38–41 help the hydrogen–propane–air mixtures get ignite with lesser hydrogen.

Fig. 11 The concentration of H₂ required in micro-combustors for ignition in hydrogen-assisted co-feed mode for different inlet velocities and wall thermal conductivities. φ (C₃H₈) = 0.7, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

4.1.2 Transient response and ignition characteristics. The equivalence ratio φ of premixed propane–air mixtures is kept fixed (0.7) and the concentration (on a molar basis) of H₂ required is 0.0005 excess of the minimum H₂ concentration required for ignition. In the co-feed ignition mode, the transient responses of bulk gas phase temperature, interior wall temperature and C₃H₈ conversion for three different values of wall thermal conductivity are shown in Fig. 12–14, respectively. As before, H₂ combustion ignites in approximately 1.0–2.0 s and directly reaches its steady state profile (magenta line: H₂ in Fig. 14). For most of the simulation time, note that the H₂ conversion profiles do not show any noticeable change. At various different times, the axial variation in C₃H₈ conversion is plotted in Fig. 14. The ignition time (t_ign) is defined as the time taken for C₃H₈ conversion rate to reach 50% at the micro-combustor exit. As the wall thermal conductivity is increased, the ignition time increases as observed in Fig. 14: from 98 s for insulating walls (0.5 W m⁻¹ K⁻¹), to 202 s for moderately thermal conducting wall (20 W m⁻¹ K⁻¹), and 448 s for highly thermal conducting walls (200 W m⁻¹ K⁻¹). The above-mentioned trend can be attributed to heat localization in the wall materials with low thermal conductivity, which leads to hot spot formation and faster ignition. In addition, in the co-feed ignition mode, the front-end ignition is observed in all cases. Consequently, the bulk gas phase and interior wall temperatures (Fig. 12 and 13) reach their steady state values quickly once C₃H₈ is ignited. Therefore, in the co-feed ignition mode, the combustion characteristics of hydrogen-assisted propane qualitatively resemble the selectively preheating initial portion of the combustion chamber wall.^3,4

Fig. 12 Transient response of bulk gas phase temperature profiles for ignition in hydrogen-assisted co-feed mode for different wall thermal conductivities. φ (C₃H₈) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 13 Transient response of interior wall temperature profiles for ignition in hydrogen-assisted co-feed mode for different wall thermal conductivities. φ (C₃H₈) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 14 Transient response of C₃H₈ conversion profiles for ignition in hydrogen-assisted co-feed mode for different wall thermal conductivities. φ (C₃H₈) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

As observed in Fig. 13, for low thermal conductivity walls (0.5 W m⁻¹ K⁻¹), the interior wall temperatures first drop, then increase and then drop again at steady state. The above-mentioned trend is mostly because H₂ combustion zone is located significantly upstream compared to C₃H₈. In the region between them, the interior wall temperatures drop slightly by reason of heat losses to the surroundings.

The gas phase temperature and C₃H₈ conversion vs. the cumulative time at the micro-combustor exit for different wall conductivities are shown in Fig. 15. As observed, the triangles represent the ignition time t_ign at which C₃H₈ is ignited, and the circles represent the steady state time t_ss at which C₃H₈ combustion taken to reach steady state.

Fig. 15 Exit temperature and C₃H₈ conversion profiles for ignition in hydrogen-assisted co-feed mode for different wall thermal conductivities. φ (C₃H₈) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹. The triangles represent the ignition time t_ign, and the circles represent the steady state time t_ss.

In general, under lean combustion conditions and co-feed mode, the premixed hydrogen–propane–air mixtures can lead to micro-combustor ignition without the need for external heating. H₂ ignites instantaneously, and followed by C₃H₈ ignition in micro-combustors after C₃H₈ ignition temperature is reached. The concentration (on a molar basis) of H₂ required for C₃H₈ ignition varies in the range of 0.008–0.028, and decreases for: lower wall thermal conductivity, higher inlet velocity, and higher equivalence ratio of propane–air mixtures.

4.2. Switched from hydrogen–air to propane–air mixtures (sequential feed mode)

4.2.1 H₂ requirement for C₃H₈ ignition. In the sequential feed mode, effect of equivalence ratio φ_C3H8 on the minimum equivalence ratio of H₂ required for C₃H₈ ignition for different inlet velocities is shown in Fig. 16. As the equivalence ratio φ_C3H8 is increased, the minimum concentration of H₂ required for C₃H₈ ignition decreases. For higher inlet velocity, the concentration of H₂ required for C₃H₈ ignition is expected to be lower in agreement with the previous section. The region below each curve represents that the H₂ concentration is insufficient for C₃H₈ ignition. While the above-mentioned behavior is existent at higher C₃H₈ equivalence ratio (φ_C3H8 > 0.73), the situation is entirely different at lower C₃H₈ equivalence ratio (φ_C3H8 < 0.64). Note that the wall temperature is still higher as the velocity is increased, keeping equivalence ratio constant. However, the location of the combustion zone is pushed downstream at higher inlet velocities. This prevents stabilization of propane combustion when we switch over from hydrogen–air to propane–air inlet. Hence, we see a reversal in the trend at lower propane equivalence ratio.

Fig. 16 The minimum equivalence ratio of H₂ required for the C₃H₈ ignition of a particular equivalence ratio φ_C3H8 for different inlet velocities.

In the sequential feed mode, effect of wall thermal conductivity on exit wall temperature and the minimum concentration of H₂ required for C₃H₈ ignition for different inlet velocities are shown in Fig. 17. The exit wall temperature shown in Fig. 17 is the minimum wall temperature that should be attained before switching from hydrogen–air to propane–air mixtures to guarantee stable C₃H₈ ignition through the parametric study.

Fig. 17 Effect of wall thermal conductivity on exit wall temperature and the minimum concentration of H₂ required for C₃H₈ ignition for different inlet velocities.

In the sequential feed mode, as wall thermal conductivity is decreased, the concentration of H₂ required decreases consistently with the co-feed mode results. However, above-mentioned variation trend is not monotonic unlike the co-feed mode. Such as inlet velocity of 4 m s⁻¹, as the wall thermal conductivity is decreased until 6 W m⁻¹ K⁻¹, the minimum concentration of H₂ required decreases monotonically. However, as the wall thermal conductivity is further decreased to 0.5 W m⁻¹ K⁻¹, the minimum concentration of H₂ required increases slightly and then continue to decrease. Similar variation trends can be observed for the inlet velocities of 2 and 8 m s⁻¹ as well. The main reason for the non-monotonic behavior depends on two competitive factors: the rate of the heat (wall) loss to the ambient air vs. the rate of the heat (wall and exhaust gas of burned hydrogen) transfer to the unburned fuel (propane) from hydrogen–air switch to propane–air for different wall thermal conductivity.

4.2.2 Transient response and ignition characteristics. In the sequential feed mode, the effect of wall thermal conductivity on the H₂ cut-off time, C₃H₈ ignition time, 99% C₃H₈ conversion and steady state time are shown in Fig. 18. For three different wall thermal conductivities (0.5, 20 and 200 W m⁻¹ K⁻¹), the temporal variations in bulk gas phase temperature, interior wall temperature and C₃H₈ conversion profiles are shown in Fig. 19–21. The flow is switched from hydrogen–air to propane–air mixtures as the exit wall temperatures in micro-combustor exceed the corresponding temperatures shown in Fig. 17(b). The switch is represented as circles in Fig. 18 and by blue lines in Fig. 19. As the wall thermal conductivity is increased, the H₂ cut-off times do not vary significantly (40 s ≤ t_cut-off ≤ 60 s). For all cases, C₃H₈ ignition times require less than 10 s and 99% C₃H₈ conversion is reached (as observed in Fig. 18) in a short time thereafter once the flow of propane–air mixtures is started. However, compared with lower thermal conductivity walls, the times taken to reach steady state are significantly higher for highly thermal conductivity walls. For low and moderate wall thermal conductivity values (0.5 and 20 W m⁻¹ K⁻¹) as observed in Fig. 19–21, the ignition characteristics are similar to partially preheating the initial segment of the micro-combustor. However, for high wall thermal conductivity values (200 W m⁻¹ K⁻¹), the ignition characteristics are close to completely heating the micro-combustor wall, and the back-end ignition (similar to fully preheated wall case⁴²) can be observed.

Fig. 18 Effect of wall thermal conductivity on H₂ cut-off time, C₃H₈ ignition time, 99% C₃H₈ conversion and steady state time in the sequential feed mode. φ (H₂) = 0.148, φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 19 Transient response of bulk gas phase temperature profiles in the sequential feed mode for different wall thermal conductivities. φ (H₂) = 0.148, φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 20 Transient response of interior wall temperature profiles in the sequential feed mode for different wall thermal conductivities. φ (H₂) = 0.148, φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

Fig. 21 Transient response of C₃H₈ conversion profiles in the sequential feed mode for different wall thermal conductivities. φ (H₂) = 0.148, φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

In the sequential feed mode, the cumulative C₃H₈ emissions and exit gas temperature for different wall thermal conductivities are shown in Fig. 22. Initially, there are no C₃H₈ emissions until C₃H₈ flow is started as observed in Fig. 22(a). The C₃H₈ emissions with sharp jump occurs between the switch time and 99% C₃H₈ conversion time (represented as circles in Fig. 22(a)). Moreover, as the inlet flow is switched from lean hydrogen–air (φ_H2 = 0.148) to propane–air (φ_C3H8 = 0.7) mixtures, the exit gas temperature increases rapidly. For high thermal conductivity walls (200 W m⁻¹ K⁻¹), net C₃H₈ emissions (22.84 mg) at the end of 200 s increased by 3 times for low thermal conductivity walls (5.54 mg (0.5 W m⁻¹ K⁻¹) and 5.58 mg (20 W m⁻¹ K⁻¹)).

Fig. 22 Cumulative C₃H₈ emissions and exit gas temperature in the sequential feed mode for different wall thermal conductivities. φ (H₂) = 0.148, φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹. Circles represent the time taken to reach 99% C₃H₈ conversion in panel (a).

The cumulative amount of H₂ and the time required for C₃H₈ catalytic ignition as a function of H₂ concentration in the co-feed and sequential feed modes are shown in Fig. 23. The cumulative amount of H₂ is defined as the total H₂ amount supplied from the beginning to characteristic ignition temperature.¹⁹ As observed in Fig. 23, the cumulative amount of H₂ and the time required for C₃H₈ catalytic ignition both decrease rapidly with increasing H₂ concentration. For the 0.02 H₂ concentration (on a molar basis), the times required to reach the ignition of C₃H₈ are 9.6 (co-feed mode) and 10.2 (sequential feed mode) times longer than those of mixtures containing 0.06 hydrogen. Similar to aforementioned behavior, for the 0.02 H₂ concentration, the cumulative amounts of H₂ are 3.2 (co-feed mode) and 3.4 (sequential feed mode) times than those of mixtures containing 0.06 hydrogen. Therefore, from both the ignition time viewpoint and minimum utilization of H₂, flows with higher H₂ concentration are desirable. These results are consistent with the conclusions from the previous works of Norton and Vlachos,¹⁷ and Zhong et al.^19–21 who used n-butane as fuel.

Fig. 23 The cumulative amount of hydrogen and the time required for propane catalytic ignition as a function of hydrogen concentration in the co-feed and sequential feed modes. Co-feed mode: φ (C₃H₈) = 0.7, sequential feed mode: φ (C₃H₈, after H₂ cut-off) = 0.7, u₀ = 2 m s⁻¹, h = 20 W m⁻² K⁻¹, and k_s = 20 W m⁻¹ K⁻¹.

In the co-feed mode, although C₃H₈ inhibits the H₂ ignition at short times,¹⁷ C₃H₈ ignites as soon as the ignition temperature is reached, leading to a faster heat up (shorter ignition time as observed in Fig. 23) of the micro-combustor as compared to the sequential feed mode. Through external controls and optimization, the ignition time for introducing propane in the sequential feed mode may be improved. However, the co-feed mode (simultaneous starting of flows) is simple, easy to operate, and has the additional benefit of heat release from both fuels (H₂ and C₃H₈). Therefore, the co-feed mode appears to be preferred. These results are consistent with the conclusions from the previous work of Norton and Vlachos,¹⁷ who compared the co-feed mode with the swapping mode.

5. Conclusions

In this study, the hydrogen self-ignition and hydrogen-assisted ignition of propane–air mixtures from ambient cold-start conditions were investigated numerically in Pt/γ-Al₂O₃ catalytic micro-combustors. For the steady and transient state, the micro-scale combustion and self-ignition characteristics of lean hydrogen–air mixtures were presented, and the hydrogen-assisted combustion of propane–air mixtures was investigated numerically in the co-feed mode and the sequential feed mode. The following conclusions were obtained from this micro-scale combustion characteristics study.

(1) Combustion and self-ignition characteristics of hydrogen in Pt/γ-Al₂O₃ catalytic micro-combustors.

• H₂ conversion reaches the steady state in less than 1.0 s and front-end ignition is observed.

• Compared with the diffusion and reaction processes, the large thermal inertia of the micro-combustor solid structure leads to slow temperature dynamics, and transient response is dominated by the thermal inertia.

• The time required to reach steady state decreases with increasing inlet velocity and decreasing wall thermal conductivity.

• The heat localization in poorly conducting walls leads to fast ignition and shorter steady state time.

(2) Micro-scale ignition characteristics of hydrogen-assisted propane in the co-feed mode (premixed hydrogen–propane–air mixtures).

• The concentration (on a molar basis) of H₂ required for C₃H₈ ignition varies from 0.008 to 0.028.

• The concentration of H₂ required for C₃H₈ ignition increased with increasing wall thermal conductivity, decreasing inlet velocity, and decreasing inlet equivalence ratio of propane–air mixtures.

• The combustion characteristics of hydrogen-assisted propane qualitatively resemble the selectively preheating initial portion of the combustion chamber wall.

• H₂ ignites instantaneously, and followed by C₃H₈ ignition in micro-combustors after C₃H₈ ignition temperature is reached.

(3) Micro-scale ignition characteristics of hydrogen-assisted propane in the sequential feed mode (switched from hydrogen–air to propane–air mixtures).

• The wall thermal conductivity, inlet velocity, and inlet equivalence ratio of propane–air mixtures have significant effect on the concentration of H₂ required for C₃H₈ ignition in the micro-combustors. In general, the concentration of H₂ required for C₃H₈ ignition increased with increasing wall thermal conductivity, decreasing inlet velocity, and decreasing inlet equivalence ratio of propane–air mixtures. While the above-mentioned behavior is existent at higher wall thermal conductivity (k_s > 30 W m⁻¹ K⁻¹) for wall thermal conductivity and inlet velocity.

• The time taken to reach steady state, the H₂ cut-off time, the C₃H₈ ignition time and the cumulative C₃H₈ emissions increased with increasing wall thermal conductivity.

• For low and moderate wall thermal conductivity values (0.5 and 20 W m⁻¹ K⁻¹), the ignition characteristics are similar to partially preheating the initial segment of the micro-combustor. However, for high wall thermal conductivity values (200 W m⁻¹ K⁻¹), the ignition characteristics are close to completely heating the micro-combustor wall, and the back-end ignition can be observed.

• Higher hydrogen concentrations lead to minimum utilization of hydrogen and a relatively fast startup of the micro-combustor, and the co-feed mode of hydrogen-assisted ignition is a good startup strategy than the sequential feed mode.

References

1. N. S. Kaisare and D. G. Vlachos, A review on microcombustion: Fundamentals, devices and applications, Prog. Energy Combust. Sci., 2012, 38, 321–359.
2. Y. Ju and K. Maruta, Microscale combustion: Technology development and fundamental research, Prog. Energy Combust. Sci., 2011, 37, 669–715.
3. K. Maruta, Micro and mesoscale combustion, Proc. Combust. Inst., 2011, 33, 125–150.
4. D. C. Walther and J. Ahn, Advances and challenges in the development of power-generation systems at small scales, Prog. Energy Combust. Sci., 2011, 37, 583–610.
5. M. L. Gross, T. D. Hedman, S. F. Son, T. L. Jackson and M. W. Beckstead, Coupling micro and meso-scale combustion models of AP/HTPB propellants, Combust. Flame, 2013, 160, 982–992.
6. S. A. Fanaee and J. A. Esfahani, Two-dimensional analytical model of flame characteristic in catalytic micro-combustors for a hydrogen–air mixture, Int. J. Hydrogen Energy, 2014, 39, 4600–4610.
7. K. Maruta, K. Takeda, J. Ahn, K. Borer, L. Sitzki, P. D. Ronney and O. Deutschmann, Extinction limits of catalytic combustion in microchannels, Proc. Combust. Inst., 2002, 29, 957–963.
8. T. A. Wierzbicki, I. C. Lee and A. K. Gupta, Combustion of propane with Pt and Rh catalysts in a meso-scale heat recirculating combustor, Appl. Energy, 2014, 130, 350–356.
9. S. S. Sadeghi, S. Tabejamaat, M. Baigmohammadi and J. Zarvandi, An experimental study of the effects of equivalence ratio, mixture velocity and nitrogen dilution on methane/oxygen pre-mixed flame dynamics in a meso-scale reactor, Energy Convers. Manage., 2014, 81, 169–183.
10. A. Alipoor and K. Mazaheri, Studying the repetitive extinction-ignition dynamics for lean premixed hydrogen–air combustion in a heated microchannel, Energy, 2014, 73, 367–379.
11. G. P. Gauthier, G. M. G. Watson and J. M. Bergthorson, Burning rates and temperatures of flames in excess-enthalpy burners: A numerical study of flame propagation in small heat-recirculating tubes, Combust. Flame, 2014, 161, 2348–2360.
12. V. N. Kurdyumov and M. Matalon, Analysis of an idealized heat-recirculating microcombustor, Proc. Combust. Inst., 2011, 33, 3275–3284.
13. H. Y. Shih and Y. C. Huang, Thermal design and model analysis of the Swiss-roll recuperator for an innovative micro gas turbine, Appl. Therm. Eng., 2009, 29, 1493–1499.
14. B. J. Tsai and Y. L. Wang, A novel Swiss-Roll recuperator for the microturbine engine, Appl. Therm. Eng., 2009, 29, 216–223.
15. P. D. Ronney, Analysis of non-adiabatic heat-recirculating combustors, Combust. Flame, 2003, 135, 421–439.
16. J. W. Döbereiner, Neu entdeckte merkwürdige Eigenschaften des Platinsuboxyds, des oxydirten Schwefel-Platins und des metallischen Platinstaubes (Newly discovered remarkable properties of platinum suboxide, oxidized platinum sulfide and metallic platinum dust), J. Fortschr. Chem., Phys., 1823, 38, 321–326.
17. D. G. Norton and D. G. Vlachos, Hydrogen assisted self-ignition of propane/air mixtures in catalytic microburners, Proc. Combust. Inst., 2005, 30, 2473–2480.
18. Y. Yan, W. Pan, L. Zhang, W. Tang, Z. Yang, Q. Tang and J. Lin, Numerical study on combustion characteristics of hydrogen addition into methane–air mixture, Int. J. Hydrogen Energy, 2013, 38, 13463–13470.
19. B. J. Zhong and F. Yang, Characteristics of hydrogen-assisted catalytic ignition of n-butane/air mixtures, Int. J. Hydrogen Energy, 2012, 37, 8716–8723.
20. B.-J. Zhong, Y.-W. Yu and P.-F. Xiong, Experimental study on catalytic ignition of n-C₄H₁₀/H₂/air mixtures, CIESC J., 2014, 65, 4363–4367.
21. B.-J. Zhong, P.-F. Xiong, F. Yang and Y.-W. Yu, Thermal and chemical effects of hydrogen in catalytic ignition of n-butane/air mixture, Acta Phys.-Chim. Sin., 2013, 29, 897–902.
22. O. Deutschmann, L. I. Maier, U. Riedel, A. H. Stroemman and R. W. Dibble, Hydrogen assisted catalytic combustion of methane on platinum, Catal. Today, 2000, 59, 141–150.
23. P.-A. Bui, E. A. Wilder, D. G. Vlachos and P. R. Westmoreland, Hierarchical reduced models for catalytic combustion: H₂/Air mixtures near platinum surfaces, Combust. Sci. Technol., 1997, 129, 243–275.
24. D. G. Norton, E. D. Wetzel and D. G. Vlachos, Fabrication of single-channel catalytic microburners: effect of confinement on the oxidation of hydrogen/air mixtures, Ind. Eng. Chem. Res., 2004, 43, 4833–4840.
25. J. Mantzaras, Catalytic combustion of syngas, Combust. Sci. Technol., 2008, 180, 1137–1168.
26. Y. Men, G. Kolb, R. Zapf, H. Pennemann and V. Hessel, Total combustion of propane in a catalytic microchannel combustor, Chem. Eng. Res. Des., 2009, 87, 91–96.
27. B. J. Zhong, Q. T. Yang and F. Yang, Hydrogen-assisted catalytic ignition characteristics of different fuels, Combust. Flame, 2010, 157, 2005–2007.
28. U. Rana, S. Chakraborty and S. K. Som, Thermodynamics of premixed combustion in a heat recirculating micro combustor, Energy, 2014, 68, 510–518.
29. W. Yang, C. Deng, J. Zhou, J. Liu, Y. Wang and K. Cen, Experimental and numerical investigations of hydrogen–air premixed combustion in a converging-diverging micro tube, Int. J. Hydrogen Energy, 2014, 39, 3469–3476.
30. M. Schultze, J. Mantzaras, R. Bombach and K. Boulouchos, An experimental and numerical investigation of the hetero-/homogeneous combustion of fuel-rich hydrogen/air mixtures over platinum, Proc. Combust. Inst., 2013, 34, 2269–2277.
31. I. Sen and A. K. Avci, Simulation of exhaust gas reforming of propane in a heat exchange integrated microchannel reactor, Int. J. Hydrogen Energy, 2014, 39, 844–852.
32. T. Kamada, H. Nakamura, T. Tezuka, S. Hasegawa and K. Maruta, Study on combustion and ignition characteristics of natural gas components in a micro flow reactor with a controlled temperature profile, Combust. Flame, 2014, 161, 37–48.
33. D. G. Norton and D. G. Vlachos, A CFD study of propane/air microflame stability, Combust. Flame, 2004, 138, 97–107.
34. N. S. Kaisare, Jay H. Lee and A. G. Fedorov, Hydrogen generation in a reverse-flow microreactor: 1. Model formulation and scaling, AIChE J., 2005, 51, 2254–2264.
35. T. Kikas, I. Bardenshteyn, C. Williamson, C. Ejimofor, P. Puri and A. G. Fedorov, Hydrogen production in a reverse-flow autothermal catalytic microreactor: From evidence of performance enhancement to innovative reactor design, Ind. Eng. Chem. Res., 2003, 42, 6273–6279.
36. Y. S. Matros, Unsteady Processes in Catalytic Reactors (Studies in Surface Science and Catalysis series), Elsevier, Amsterdam, 1985.
37. G. B. Chen, C. P. Chen, C. Y. Wu and Y. C. Chao, Effects of catalytic walls on hydrogen/air combustion inside a micro-tube, Appl. Catal., A, 2007, 332, 89–97.
38. G. Bagheri, S. E. Hosseini and M. A. Wahid, Effects of bluff body shape on the flame stability in premixed micro-combustion of hydrogen–air mixture, Appl. Therm. Eng., 2014, 67, 266–272.
39. V. Shirsat and A. K. Gupta, A review of progress in heat recirculating meso-scale combustors, App. Energy, 2011, 88, 4294–4309.
40. G. A. Boyarko, C. J. Sung and S. J. Schneider, Catalyzed combustion of hydrogen–oxygen in platinum tubes for micro-propulsion applications, Proc. Combust. Inst., 2005, 30, 2481–2488.
41. J. Hua, M. Wu and K. Kumar, Numerical simulation of the combustion of hydrogen–air mixture in micro-scaled chambers. Part I: Fundamental study, Chem. Eng. Sci., 2005, 60, 3497–3506.
42. L. Petzold, Automatic selection of methods for solving stiff and nonstiff systems of ordinary differential equations, SIAM J. Sci. Stat. Comput., 1983, 4, 136–148.

---

Footnote

† Current address: School of Mechanical and Power Engineering, Henan Polytechnic University, 2000 Century Avenue, Jiaozuo, Henan, 454000, P.R.China.

---