Effect of channel geometry and porous coverage on flame acceleration in hydrogen–air mixture

Abstract

Flame propagation in a hydrogen–air mixture in the presence of porous materials was investigated experimentally in channels with different dimensions and cross–sections. In this study, experiments were performed in a rectangular channel with one or two walls covered with porous material to study flame propagation in stoichiometric hydrogen–air mixtures at room temperature and atmospheric pressure. Depending on the channel configuration, the porous coating of the internal walls ranged from 1/4 to 1/2 of the channel area. Four types of polyurethane foam with a number of pores per inch (PPI) ranging from 10 to 80 were used to cover the channel walls. Flame propagation was visualized using a Schlieren device and high-speed camera. The largest flame acceleration in the porous channel relative to the solid channel was observed in the 20 × 20 mm channel. The ratio of the velocities in the porous channel to the velocity in the solid channels was 6–7 for the porous material with the largest (2.5 mm) pores. In the case of a 10 × 10 mm channel, the flame velocity in the porous channel was higher than the flame velocity in the solid channel after 350 mm only when using porous coatings with 2.5 mm and 1.3 mm pores. When using a porous coating with smaller pores the flame velocity was lower than in the solid channel. Schlieren images show different stages of flame propagation from a turbulent flame to a supersonic flame with shock waves.

Introduction

One of the modern energy safety concerns is ensuring explosion safety when working with flammable gases. The most urgent problem is hydrogen safety due to the fact that hydrogen has wide concentration limits of combustion and low ignition energy (Molnarne and Schroeder, 2019). The solution to the problem of preventing flame acceleration and detonation formation is relevant for devices using hydrogen and hydrogen storing facilities.

The use of porous coatings and channel obstacles has become an important topic in the combustion and detonation research during the last decades. The use of porous elements makes it possible to solve the problem of increasing explosion safety.

It is worth noting a number of works showing the efficiency of using porous coatings for detonation decay. The use of various materials, structures and various gases made it possible to establish a number of patterns for this type of combustion. In the works by Teodorczyk and Lee (1995) and Bivol et al. (2016) it was shown that the detonation decayed into a shock wave and a flame front, which moved at 0.5 of the Chapman–Jouguet detonation velocity. In the works by Guo et al. (2002) and Golovastov et al. (2019) the effect of porous and perforated coatings with high thermal conductivities on the detonation decay was shown. It was found that steel wool was the most effective at attenuating the detonation wave. The role of the channel coverage ration for detonation suppression was shown by Rao et al. (2019). When changing coverage from 1/4 to 3/4 of the channel inner area the shock wave velocity dropped from 1796 m/s to 617 m/s. Four different detonation propagation regimes were discovered after passing through a polypropylene membrane (Guan et al., 2018). With increase of the membrane thickness the detonation propagation regime changed from velocity deficit mode to failure mode.

A multiparametric study of the effect of porous coatings on the unsteady propagation of the flame is a complex problem, which is complicated by the geometric formulation, boundary conditions and by the presence of turbulence in the flow of reacting gases caused by these porous coatings.

Previous studies on the deflagration to detonation transition (DDT) showed the dependence of DDT distance on the channel diameter (Baumann et al., 1961, Cubbage, 1963). More recently it was discovered, that the transition to detonation can proceed differently in smaller channels compared to larger channels due to viscous effects (Han et al., 2017). It was also discovered, that the relationship between channel diameter and DDT distance can change depending on the mixture composition (Li et al., 2006). For stoichiometric and fuel-rich mixtures, a sublinear relationship between DDT distance and channel diameter was observed. For fuel-lean mixtures the sublinear dependence between DDT distance and channel diameter was not observed. The acceleration of the flame in hydrogen mixtures significantly depended on the composition of the mixture and the size of the channel (Dorofeev, 2009). In 20 mm and 30 mm wide channels, the DDT distance increased linearly with the channel height at any initial pressure (Li et al., 2017). Channel geometry can also influence subsonic flame propagation (Wang et al., 2020). It was discovered that a larger increase in channel cross–section led to a larger maximum overpressure.

The presence of obstacles or irregularities in the channel can lead to the formation of turbulence ahead of the flame front. The presence of turbulence can significantly increase the reaction rate and lead to an increase in pressure and flame acceleration (Silvestrini et al., 2008). The use of porous coatings can have the same effect as the use of obstacles. Flame perturbations size in hydrogen–air mixtures was found to increase with increasing pore size of the porous material on the channel walls (Golovastov et al., 2021). The acceleration of a flame in a mixture of hydrogen and oxygen was visualized using the shadow method in a smooth channel and in a channel with surface irregularities (Maeda et al., 2019). It was found that only slow subsonic combustion was observed in the smooth channel, while irregularities in the channel led to the acceleration of the flame to detonation at a distance of 120 mm from the ignition point. It was shown that the propagation of the flame inside the porous layer is the reason for the acceleration of the flame in the channel, but the porosity did not affect the deflagration-to-detonation process (Johansen and Ciccarelli, 2008).

The use of porous coatings and obstacles can also lead to the formation of a detonation. The transition from deflagration to detonation was experimentally studied in a channel with additional tubes inside (Sun and Lu, 2020). The results showed that the critical pressure for the transition of combustion to detonation decreased when using tubes in the channel. When the flame propagated in the channel with obstacles, the reflection of shock waves from the obstacles and turbulence were detected, which led to an increase in pressure near the flame front and heating of the unburned mixture (Houim and Oran, 2017). DDT in a square channel filled with obstacles was experimentally studied in a hydrogen–methane–air mixture (Wang et al., 2018). It was concluded, that square obstacles led to more significant detonation facilitation than fence-type obstacles. The influence of orifice plate shapes and spacing was experimentally studied (Wang et al., 2019). The orifice plate shape had no effect on the detonation velocity deficit in respect to Chapman–Jouguet detonation velocity if the effective diameter stayed the same. The effect of the channel height on flame acceleration was considered in a channel with a porous coating (Ciccarelli et al., 2009). It was found that the transition from deflagration to detonation in a channel with a height of 74 mm occurred later than in a channel with a height of 38 mm. The detonation velocity deficit was lower in the 74 mm high channel.

In certain cases, the opposite effect of suppressing combustion with porous materials was observed. When three different distances from porous copper to the ignition point were considered, the combustion inhibition effect was most pronounced when the porous material was located close to the ignition point (Shao et al., 2020). Multilayer obstacles made of steel with different hole sizes were used to extinguish a hydrogen–air flame. It was found that the flame was extinguished using 45 and 50 layers of mesh 60, 80 and 100. However, with mesh 40, even 50 layers were not enough to extinguish the combustion (Jin et al., 2017). The flame quenching performance of wire mesh was compared for fuel-rich and fuel-lean mixtures (Jin et al., 2021). The flame suppression effect was more pronounced in fuel-rich mixtures.

A number of obstacles with different thicknesses and a porous material with different numbers of pores were used to study flame propagation. It was found that placing obstacles before the porous material caused the flame to accelerate. The flame extinguishing effect was enhanced with a decrease in the pore size of the porous material and with an increase in its thickness (Wen et al., 2013). It was found that in a channel with a porous coating on the wall, a smaller increase in pressure is observed than in a solid channel (Chen et al., 2017).

It should be noted that under certain conditions, criteria have been established to determine the propagation mode of the flame in a space with a porous coating, including the properties of the porous materials. The works (Bivol and Golovastov, 2020, Lyamin and Pinaev, 1986) established that a Peclet number equal to 50–60 allowed the flame to propagate inside the porous material and accelerate the flame when the walls were covered with porous material.

The effect of the hole size of round obstacles on the propagation of the flame front in a hydrogen–air mixture was investigated. It has been found that smaller holes result in more turbulence and more flame acceleration when passing through obstacles (Liu et al., 2019). With an increase in the initial pressure and enrichment of the mixture, the degree of flame acceleration with the help of obstacles also increased. DDT in hydrogen–air in a hydrogen–oxygen mixture was experimentally studied in a channel with two obstacles of different blockage ratio (Ahumada et al., 2020). It was discovered that fastest transition to detonation was recorded when using obstacles with a sharp increase in blockage ratio. Flame acceleration in hydrogen–air mixtures was experimentally studied in flat channels. It was found that the acceleration of the flame to the speed of sound depended on the composition of the mixture and the thickness of the channel (Grune et al., 2013). The propagation of a flame in a volume with a perforated plate was experimentally studied. Six different modes of flame propagation were found, depending on the composition of the combustible mixture and the size of the holes in the plate (Zhou et al., 2018).

The size of the channel and the degree of coverage can affect the propagation of the flame front. Changing the degree of coverage and channel dimensions it is possible to find conditions for flame acceleration and deceleration. In our work channels with a width of 10 mm and 20 mm and porous coverage of 1/4, 1/3 and 1/2 were used. The aim of this work was to study the effect of the channel coverage with a porous material and the channel dimensions on flame acceleration in a stoichiometric mixture of hydrogen with air. The results can help develop methods for passive detonation suppression in tubes or hydrogen cells.