Critical review of Molkov's phenomenological model and variable stretch/turbulence function. (September 2016)

• Record Type: Journal Article
• Authors: Hernandez, Francisco; Abdel-jawad, Madhat; Hao, Hong
• Is Part Of: Journal of loss prevention in the process industries. Volume 43(2016:Sep.)
• Page Start: 225
• Page End: 241
• Publication Date: 2016-09
• Subjects: Gas explosion -- Ventilation -- Molkov's model -- Discharge sub-model -- Simplified multiple equations' inverse problem -- FLACS; Chemical industries -- Safety measures -- Periodicals; 660.2804
• DOI: 10.1016/j.jlp.2016.05.019
• Languages: English
• ISSNs: 0950-4230

Abstract

Abstract: A critical review of Molkov's phenomenological model including the model's assumptions and sub-models is presented in this article. First, the effect of an incorrect discharge sub-model is studied and it is concluded that the choice of discharge sub-model is crucial because it can substantially modify the turbulence/stretch function, the mass distributions and the flame position. Therefore, a discharge sub-model should only be chosen when the sub-model is in agreement with the flame position and the residual unburnt gas mass reaches a reasonable value. Second, the equivalent flame radius and the apparent flame velocity (computed internally by Molkov's model) are found to depend on three different effects: the free flameball expansion, the adiabatic compression/extension and the venting process. Third, the interpretation of the turbulence function should account for the effect of the propagation mode and the spatial variation of the local flame speed. Fourth, the jet effect model related to hinge panels can be improved; therefore, a new model is presented. Fifth, the universal correlation and the two-lumped-parameter model are studied. Despite the high correlation reported in previous publications, it is concluded that the two-lumped-model has significant limitations and should be improved if a variable stretch/turbulence function is utilized, which will require the utilization of "usual" discharge coefficients. Finally, it is shown that the inverse problem with a variable stretch/turbulence function and a reasonable discharge coefficient can be utilized to accurately backfit experimental data. Highlights: The selection of discharge sub-model is analyzed. The equivalent flame radius is described by the three effects. The jet effect model related to hinge panels is improved. the two-lumped-model involves significant limitations. A variable turbulence function and a usual discharge coefficient should be used.