Auto-ignition delay times and detonation cell size of hydrogen-nitrous oxide-argon mixtures
Rémy Mével,Fabien Lafosse,Laurent Catoire,Nabiha Chaumeix,Gabrielle Dupré
2007-01-01
Abstract:Silicon containing gaseous compounds, such as silane or chlorosilane, are of large interest in some kinds of industries such as semiconductor and ceramic material production since they can be used in order to produce SiC, Si3N4 or SiO2 powders [1]. Using thermal or plasma enhanced chemical vapour deposition processes, silicon dioxide solid films can be formed as protective layers and electrical insulators [2]. As a consequence of the necessary conditions for the thermal processes, that are close to those used for low pressure combustion reactions, a risk of ignition and explosion has not to be excluded [2]. Among all silicon based compounds, silane seems to be the most hazardous. Indeed, it presents a very wide flammability domain, a very large flame temperature range, a pyrophoric behaviour for concentration above 4.5 mol % in air, and very high pressure rise rate and burning velocity [3]. Although Hirano [4] reports an explosion that has been attributed to silane-nitrous oxide mixtures, very few papers deal with explosives properties of this system. However, Horiguchi et al. [5] have shown that, at room temperature and atmospheric pressure, these mixtures are stable but present a wide flammability domain since it ranges from 1.90 to 87.1 mol % SiH4. An important behaviour of silane is its decomposition mechanism. First, it must be noticed that this decomposition reaction is fast for temperatures as low as 1200 K, whereas the one of nitrous oxide remains slow at this temperature [6]. Moreover, many kinetic studies have shown that silane decomposition leads to the release of molecular hydrogen [7]. It follows that the H2-N2O chemistry might play an important role in the SiH4-N2O system. Some studies have been carried out on induction delay times of these mixtures [8], essentially at low pressure. Besides, the detonation sensitivity of H2-N2O-(N2) mixtures with respect to equivalence ratio Φ, to diluent percentage and to initial pressure has been investigated [8-9]. As part of this study, several available kinetic models have been tested. The so-called “modified” Miller and Bowman model has been reported as the best one chosen to model induction times of H2-N2O-Diluent mixtures [8]. The aim of the present study is to investigate the high pressure behaviour of H2-N2O-Ar mixtures via a characteristic time of reaction (τc) defined thanks to the OH radicals UV emission. Besides, using the “modified” Miller and Bowman model and an OH* sub-mechanism, the experimental data have been modelled. Finally, using the resulting mechanism and detonation cell width (or size) data previously determined at Caltech [8-9], an attempt has been made to correlate the cell size-induction distance ratio with initial conditions.