Effects of heat transfer on flame stability limits in a planar micro-combustor partially filled with porous medium

Abstract

Inserting porous medium into micro-combustors is able to further enhance heat recirculation, thus favoring the extension of flame stability limits. Both experimental and numerical studies have shown that extended standing-wave combustion regimes exist in micro-combustors filled with porous medium. However, the underlying mechanism that dictates the critical flame stability limits is not well understood. As such, numerical simulations are conducted for a planar micro-combustor partially filled with porous medium in order to quantify heat transfer and to analyze its effects on the critical conditions under which flames will break the upper or lower boundaries of the porous medium. Based on the proposed definitions of the preheat zone and the heat loss zone, preheating and heat loss occurring inside the porous micro-combustor are quantified. It is shown that neither preheating nor heat loss alone is sufficient to determine the flame stability limits. Therefore, their relative importance is considered by a ratio (R_p-hl) which measures the net heat gain inside the porous zone. Making use of the correlations of R_p-hl with the flow velocity and the equivalence ratio, a flow velocity range is obtained beyond which the flame cannot be stabilized within the porous medium, regardless of the equivalence ratio used. A parametric study is subsequently carried out to examine the effects of two important parameters, they are, the thermal conductivity and the porosity of the porous medium, with the results briefly analyzed. In principal, the approach presented in this paper could be readily applied to other configurations of micro-combustors as well. By incorporating pore-scale flow, heat and mass transfer details into the numerical model, the results are expected to be more quantitatively accurate.

Introduction

Since Epstein and Senturia proposed the concept of ‘micro machinery’ in 1991, combustion-based micro power devices have attracted remarkable attention of scientists and engineers all over the world [1]. Various types of micro power devices have been prototyped, developed and tested [2]. As the key component of the above-mentioned micro power devices, the micro-combustor plays a key role in determining the overall efficiency and system performance. The reduced physical dimensions of the micro-combustor give rise to shorter residence times and higher heat loss fractions, and therefore flame stability and flammability limits are the main challenges to the design and operation of micro-combustors. Achieving self-sustained and stabilized flames in micro-combustors is the primary requirement for downstream power generation or conversion devices, for example, the micro thermophotovoltaic (TPV) [3] or the thermoelectric (TE) systems [4].

Theoretical analyses [5], [6] indicate that heat recirculation is crucial to sustaining stabilized combustion in micro-combustors. A good example is the Swill-roll micro-combustors [7], [8] which allow the exhaust gases to transfer heat to the unburned mixture through a non-contact counter-flow structure. In the meanwhile, heat recirculation through the solid matrix of the porous medium, commonly used in filtration combustion (also known as ‘porous medium combustion’), provides a clue to further enhance heat recirculation in micro-combustors. Li et al. [9] inserted some porous medium made of stainless steel (SS) mesh into a planar micro-combustor and conducted the experiment on its performance as a flame holder. Their further studies were carried out on the flame stability limits of such a configuration under different filling conditions [10], [11]. Unlike its macro-scale counterpart [12], [13], [14], filtration combustion in micro-combustors exhibits a wider range of stability limits. Yang et al. [15] and Fursenko et al. [16] studied filtration combustion of premixed CH₄/air in high-porosity micro-fibrous porous medium. Interestingly, they found a regime within which standing-wave combustion modes were observed. Similar results were obtained by Wan et al. [17] in both their experiment and numerical simulations.

When a micro-combustor is fully or partially filled with porous medium, besides the combustor wall, a new pathway for heat recirculation emerges, that is, the solid matrix. In our previous study [18], the detailed methodology to quantify the two pathways of heat recirculation was developed. However, it is realized that heat recirculation alone is insufficient to understand the flame stability limits in a porous micro-combustor. Another important factor, that is, heat loss in the flame zone, should be considered in order to reveal the underlying mechanism. Thus, the primary objective of the present study is to identify the effects of heat transfer on the flame stability limits in porous micro-combustors. A planar micro-combustor partially filled with porous medium is chosen to be experimented and modelled, because this configuration has been studied extensively [9], [10], [11], [18], [19]. It has the advantages of controllable flame positions and lower flow resistances. The numerical model and subsequent analysis presented in this paper could be readily extended to micro-combustors fully filled porous medium, and therefore the present study is not limited by the specific configuration considered. In a recent publication, Sirotkin et al. [20] pointed out that the discrete structure of the solid matrix had some strong impact on the flow, heat and mass transfer inside the porous medium, and a continuum model could fail to capture some intrinsic features, especially when the pore size and the channel height are comparable. Having realized that problem, the continuum model is still employed in the present study mainly because of its simplicity and low computational cost. What is more, the porous structure of the folded SS mesh is unknown, making a pore-scale model unfeasible at this moment of time. As such, the approach of our modelling and analysis has a rigorous theoretical basis, while the numerical results should only be taken as qualitatively accurate.