Pressure threat to containment due to flame acceleration

This paper will discuss the results of a study of the acceleration of flames due to venting and obstacles (which could occur inside a containment building). The peak overpressures generated by accelerated flames are sufficient to threaten the integrity of some nuclear power plant containment buildings. Our study included a series of experiments conducted using mixtures of hydrogen and air, as well as the development of a numerical model of the phenomena. The experiments employed a large cylindrical combustion vessel that could be either closed or vented at the end opposite the igniter. Combustible mixtures used in the study contained either 15% or 30% hydrogen in air. Information obtained from this test series included flame trajectories, flame front profiles, and combustion-generated pressure histories. ''Orifice-plate'' baffles could be inserted into the combustion vessel to enhance the development of large-scale vortical motion, intensify the turbulence level, and accelerate the flame. A simple flame acceleration model was developed and incorporated into a computer code named FLACC. Briefly, a feedback mechanism between the combustion rate and the flame front displacement is the basis of the model. Spatial variations in pressure are not included in the model, so it is really only appropriate for simulating tests with mild flame acceleration and low flame speeds (however, we have used it to simulate all cases, and it generally performs well). Burned and unburned gases are compressed isentropically as the pressure in the vessel increases and the gases are allowed to exit the vessel for vented cases. FLACC appears to predict the flame velocities and combustion overpressures fairly well for all cases. We believe that in its current form, FLACC is suitable for providing a rough estimate of the combustion-generated overpressure that can be generated in a vented room or vessel. Comparison with data from other ''room'' geometries, obstacle forms, and combustible gases (e.g., mixtures containing hydrogen, carbon monoxide, air, and steam) would be required to make the code a generally applicable tool for nuclear plant safety evaluations. 17 refs, 7 figs, 1 tab