No abstract

Results of a series of tests on the deflagration of methane-air mixtures in a large vented enclosure are presented. Experiments were made in FM Global's 63.7 m 3 chamber. The chamber was 4.6 x 4.6 x 3.0 m with a vent opening on one side. Vent areas of either 2.7 or 5.4 m 2 were used. Tests were performed with ignition either at the center of the chamber or at the center of the wall opposite the vent. Methane-air mixtures with methane concentrations close to 9.5% vol. were used in the tests. Pressure data, as function of time, and flame time-of-arrival data were obtained both inside and outside the chamber near the vent. Detailed experimental data is used in the paper to test a three-dimensional gasdynamic model for the simulation of gaseous combustion in vented enclosures. The model is based on a Large Eddy Simulation (LES) solver created using the OpenFOAM CFD toolbox using sub-grid turbulence and flame wrinkling models. Results from the calculations are compared with the experimental data. The capabilities and deficiencies of the model are discussed.

Research was carried out to develop improved protection guidelines for silane handling systems through enhanced understanding of the behavior of releases of this pyrophoric gas. The approach involved addressing three aspects of the problem: the prompt ignition behavior of silane; the reactivity characteristics of quiescent silane/air mixtures; and the rates of reaction of silane leaked into enclosures with and without explosion venting, in the presence of ventilation airflow. Afirst conclusion, reached from tests in a ventilated cabinet, was that, contrary to prevailing belieJ the ventilation jlow has no measurable effect on the prompt ignition of the release. From experiments in a 5. 1-liter (31 l-in.3j sphere it was found that silane/air mixtures of concentrations between 1.4 and 4.1% (by volume) are e x p b sive but stable. In this case, piloted ignition tests yielded laminar burning velocities up to 5 m/s (1000fl/min). Mixtures between 4.5 and 38% (the maximum reached in the testd were found to be metastable, and would undergo spontaneous ignition after a delay ranging from 15 to 120 seconds, with the shorter values corresponding to higher silane concentrations. Experiments were also performed in a 0. 645-m3 (22.8$t3j vessel, both with and without explosion venting, to measure the rates of enelgy release associated with impulsivelystarted silane leaks from 1/8 and 1/4-in. (3.2 and 6.4mmj lines. A method for the prediction of the venting requirements of partial-volume deflagrations (PVDj was evolved into a tool to quantify the pressure rise from ignition of silane leak in enclosures. These results represent a significant step toward updating existing design recommendations which prescribe ventilation requirements that are based on outdated and, in some instances, m isinte p e t e d data.

The paper addresses one aspect of the definition of the explosion hazard from flammable liquid spills or slow heavy vapor releases in enclosures. In this type of accidents, the explosive mixture is typically confined to a layer near the floor. No methods are currently available for the sizing of explosion vents in these partial‐volume deflagrations. Resolution of the issue was sought through tests carried out in a 63.7 m3 (2250 ft3) chamber with stratified mixtures of propane in air. The layers were formed by slowly injecting propane at the chamber floor through diffusers, at a rate of the order of the estimated rate of vaporization of a typical solvent such as acetone. The layer composition was carefully characterized through gas concentration measurements at twelve locations, using a single continuous analyzer connected to a multiplexed sampling system. Following ignition of the layers, data were obtained on the propagation velocity of the flame and on the pressure developed by the explosion in the room both with and without venting. The results of the experiments have been correlated using an existing model, modified to account for the cylindrical geometry of the system and for the dual‐mode character of the combustion process (a premixed flame is typically followed by diffusive/ convective burning of the rich portion of the fuel/air layer). While confirming the conservatism of current design recommendations, which are based on the assumption of a full‐volume explosion, the work has brought out the fact that the contributions to pressure development from the different portions of the layer must be properly taken into account. Work is currently in progress to predict the fuel distribution in the explosive layer as a function of the parameters that define the spill (or fuel release) scenario. This step is being done through modeling of the rate of vaporization of a liquid pool and the subsequent dispersion of the vapors in the surrounding area. The model, when combined with the already‐developed treatment of the explosion problem, will provide an integrated tool for predicting the protection requirements of flammable liquid process/dispensing areas.