화학공학소재연구정보센터
Journal of Loss Prevention in The Process Industries, Vol.19, No.6, 621-629, 2006
Computer-aided modeling of the protective effect of explosion relief vents in tunnel structures
This paper aims at contributing to the efficient design of explosion protection systems against confined explosions. The issue addressed concerns the quantitative estimation of the protective effect of explosion relief vents in the case of confined explosions inside tunnels. A series of virtual experiments performed by computer simulation, revealed how the number of vents, their diameter, as well as the angle between the vents and the tunnel, influences the blast wave attenuation. The computational study was performed considering a complicated large-scale tunnel configuration with branches on its half portion. The purpose was the calculation of the attenuation effect due to the presence of vents by comparing the total explosion-specific impulse developing at antidiametric positions inside the tunnel. Simulations were carried out via a three-dimensional numerical model built in the computational fluid dynamics code CFX 5.7.1, which has been validated in previous papers against experimental overpressure histories data demonstrating reasonable performance. Computer results showed that the use of branch vents provides an effective method for shock wave attenuation following an explosion, whereas their statistical elaboration revealed that the attenuation is mainly affected by the number of vents and their diameter. In contrast, the angle between the side vents and the main tunnel appeared to slightly affect the pressure wave weakening. Eventually, the quantitative influence of the above parameters was effectively illustrated in functional diagrams, so that the total attenuation effect may be promptly estimated, if the design variables are known. In addition, two statistical models with reasonable fitting to the calculated data are proposed, which express the attenuation effect as a dependent variable of the design variables including their interactions. (C) 2006 Elsevier Ltd. All rights reserved.