Effect of the ignition position and obstacle on vented methane-air deflagration

In this study, explosion venting of front, centrally, and rear ignited 9% methane-air mixtures has been conducted in a 1-m$^3$ rectangular vessel with and without cylinders placed parallel to the venting direction. Three pressure peaks $P_1$, $P_2$, and $P_{ext}$ caused by vent failure, flame-acoustic interaction, and external explosion, respectively, can be distinguished. The pressure peak $P_1$ appears in all the tests and is insensitive to the ignition position, but the existence of obstacles increases its value. The pressure peak $P_2$ only appears in the centrally and front ignited explosions without obstacles. The pressure peak $P_{ext}$ can be observed in the rear ignition tests and is strengthened by the cylinders. The duration of the Helmholtz oscillations is longer in front ignition tests, whereas addition of cylinders had a minor effect on their frequency. This study also validates the ability of FLACS in predicting a vented methane-air explosion by comparing the simulated pressure–time histories and flame propagations with experimental results. FLACS can basically predict the shape of overpressure curves. If cylinders exist, the simulation results ensure better agreement with the experimental data because FLACS cannot simulate the flame-acoustic-interaction-induced pressure peak $P_2$. The performance of FLACS is satisfactory in rear ignition tests because it calculates $P_{ext}$ and obstacles' effect on $P_{ext}$ exactly. The flame behavior simulated by FLACS is similar to that in experiments, but the effect of the Taylor instability on the flame is not sufficiently considered.