Effects of cross-wise obstacle position on methane–air deflagration characteristics

https://doi.org/10.1016/j.jlp.2013.08.006Get rights and content

Highlights

  • Three obstacle configurations with different cross-wise positions are investigated.

  • High-speed flame images and overpressures were recorded and analyzed.

  • Cross-wise obstacle positions have significant effects on deflagration characteristics.

  • Flame speeds and overpressures for both the central and staggered configurations are greater.

  • May help designers in optimizing the internal layout of obstacles in practical processes.

Abstract

A vented chamber, with internal dimensions of 150 mm × 150 mm × 500 mm, is constructed in which the premixed methane–air deflagration flame, propagating away from the ignition source, interacts with obstacles along its path. Three obstacle configurations with different cross-wise positions are investigated. The cross-wise obstacle positions are found to have significant effects on deflagration characteristics, such as flame structure, flame front location, flame speed, and overpressure transients. The rate of flame acceleration, as the flame passes over the last obstacle, is the highest at the configuration with three centrally located obstacles, whereas the lowest is observed at the configuration with three obstacles mounted on one side of the chamber. Compared with the side configuration, the magnitude of overpressure generated increases by approximately 80% and 165% for the central and staggered configurations, respectively. Furthermore, flame propagation speeds and generated overpressures for both the central and staggered configurations are greater, which should to be avoided to reduce the risk associated with turbulent premixed deflagrations in practical processes.

Introduction

Despite the many research efforts and severe precautions made, premixed methane–air deflagration accidents still occur, which result in considerable losses in lives and property in process industries, such as chemical plants and coal mines. When deflagration occurs in confined or semi-confined regions, the flame front which moves away from the ignition source, may encounter obstacles along its path. The burning rate of a propagating flame is known to be enhanced when the flame interacts with solid obstacles (Ciccarelli et al., 2010, Ibrahim et al., 2001). The flame–obstacle interaction results in higher flame speed and overpressure, which makes premixed methane–air deflagration potentially more dangerous.

In the past three decades, a number of studies have been conducted to gain insights into the flame–obstacle interactions. Among these studies, earlier works (Dorofeev et al., 1996, Eckhoff et al., 1984, Harrison and Eyre, 1987, Hjertager et al., 1988, Moen et al., 1982) mainly focused on large-scale experiments. Moen et al. (1982) carried out methane–air explosion tests in a tube 2.5 m in diameter and 10 m long. The group found that even relatively small repeated obstacles have significant influences on the severity of explosion, generating larger explosion overpressures. Hjertager et al. (1988) developed an experimental study on large-scale flame and pressure development utilizing a 50 m3 obstructed tube, with one closed end and one open to the atmosphere, using conventional pressure and flame-arrival probes. Both pressure and flame speed data showed strong dependence on fuel-air concentrations. Harrison and Eyre (1987) studied a series of large-scale deflagrations in a 4000 m3 vessel filled with premixed methane–air and propane–air. In this research, a number of different obstacle configurations were investigated to demonstrate the effects of changes in obstacle parameters, such as height, blockage, and grid spacing, on flame speed and overpressure. However, these earlier measurements were not able to use advanced diagnostics and were thus limited in obtaining pressure-time histories and inaccurate flame speed, as large-scale experiments are very costly and often impractical.

Recently, laboratory-scale experiments have extended these measurements to monitor more accurate flame structure and flame speed. Masri, Ibrahim, Nehzat, and Green (2000) studied premixed flame propagation over three different solid obstructions, with circular, triangular, and square cross-sections covering blockage ratios ranging from about 10% to 78% in an explosion vessel (545 mm × 195 mm × 195 mm). They found that obstructions with square cross-sections resulted in the fastest flame acceleration followed by triangular and circular cross-sections. They also observed from the experimental data that flame speed increases with increasing blockage area ratio. Utilizing 20 L vented chambers, with obstructions shaped as cylinders, triangles, squares, diamonds, and plates, Ibrahim and Masri (2001) showed that the plate type obstruction resulted in the highest overpressures and that the cylindrical obstruction produced the lowest overpressure. Ciccarelli, Fowler, and Bardon (2005) investigated the effect of obstacle size and spacing on the initial stage of flame acceleration in a rough tube and found that for higher blockage ratios, flame acceleration was the greatest in a one-tube diameter spaced plate. Hall, Masri, Yaroshchyk, and Ibrahim (2009) examined the effects of the number and location of solid obstacles on the rate of turbulent premixed flame propagation. They suggested that while the peak overpressure increases with the increasing number of grids or baffle plates, a limit is reached at which the pressure starts to decrease. Furthermore, images captured by laser-induced fluorescence showed that the reaction zones become more contorted with the increasing number of baffle plates in the flame path. In similar combustion chambers, Gubba et al., 2008, Gubba et al., 2011, Masri, Al-Harbi, Meares, and Ibrahim (2012) and Masri, Ibrahim, and Cadwallader (2006) carried out measurements and large eddy simulations of propane–air mixture premixed flame, propagating past built-in solid obstructions arranged in a series of configurations with varied numbers and positions. However, the effects of cross-wise obstacle position on deflagration characteristics have been seldom investigated.

This study is a continuation of the previous works focusing on the effects of cross-wise obstacle position on methane–air deflagration characteristics, such as propagating flame structure, flame speed, and overpressure transients. Three configurations, with different cross-wise obstacle positions, in a laboratory-scale vented explosion chamber are experimentally studied.

Section snippets

Experimental details

The experimental chamber used in this study, shown schematically in Fig. 7a, has a total volume of 11.25 L, with a square cross-section of 150 mm and a height of 500 mm. The bottom end is fully closed, and the open end is sealed with a thin polyvinyl chloride (PVC) membrane to contain the premixed flammable mixture. When deflagration occurs, the thin PVC membrane is ruptured at low pressure, allowing the high-pressure gases, including unburned and burned mixtures, to escape. The chamber walls

Effects on flame structure

Fig. 3 shows the high-speed images of flame structures obtained from experimental measurements of the three configurations. Corresponding to similar flame front distances from the bottom end after ignition, eight images are selected for each configuration to study the effects of cross-wise obstacle position on flame structure. Initially, flame structures in all configurations are similar and hemispherical, as can be seen from the images at 14 ms after ignition. The flames travel relatively slow

Conclusions

In this work, we have studied the effects of cross-wise obstacle position on methane–air deflagration characteristics. Three configurations have been examined, in which three obstacles are: (1) centrally located, (2) mounted on one side of the chamber, and (3) staggered on both sides. All the cases began with a stagnant and stoichiometric methane–air mixture.

Flame front propagation and overpressure dynamics during the whole process of the deflagration are found to be different for obstacles at

Acknowledgment

This work was supported by the National Natural Science Foundation of China (Nos. 50974055 and 51176021).

Cited by (0)

View full text