N A previous investigation (4) it mas observed that the reproducibility of flash-back points for pi emixed hydrogenoxygen flames, burning at atmospheric and higher pressures, was greatly affected by the length of time in n-hich flash back was reached. Often an apparently stable flame flashed back after it had burned for several minutes, although no change in the gas flows or chamber pressure had occurred, This behavior of the flame resulted in a large scatter in the critical velocity gradient data. Similar performance TTas obtained with hydrogen-air flames. It was assumed that the erratic behavior was caused by an increase in the temperature of the burner rim. Later experiments in the investigation disclosed that conditions for flash back were reproducible when water-cooled copper tubes TTere used as burneis.I n this investigation it was attempted to elucidate the effect of burner-tip temperature on ff ash back of hydrogen-oxygen flames burning at atmospheric pressure. The effects of burner diameter, burner \Tall thickness, burner material, burner shape, mixture ratio, and ambient conditions on burner-tip temperature and critical velocity gradient n-ere examined. Over 500 tests were conducted t o determine the effects of the variables on the burner-tip temperature and velocity gradient at flash back. The principal series of tests TWS made nith burners fabiicated from copper. The general procedure of conducting the experiments consisted of determining the burner-tip temperature and velocity gradient at the burner mall a t flash back over a mixture-ratio range of from approximately 25 to SO% hydrogen by volume for each of the burners tested. Burners n-ere selected to show the influence of diameter, n-all thickness, material, and convergence of the burner tube. Edse ( 2 ) .This paper is based on iTork by Bollinger and response time. Standard conversion tables were used t o obtain the temperature. As the flash-back point was approached, the flow changes had to be made much more slowly than when further away, to ensure equilibrium. Recorder response tinica vas more than adequate for this work.Experiment,s nith small burners were conducted in the rc'strict,ed atmosphere of a closed room. Most of the combustion products were exhausted through an overhead hood and fan ajrangement. A flow diagram of thie system appears in Figure I .Hydrogen and oxygen gases were taken from standard comme~'-cia1 tanks at an initial pressure of approxiiiiately 150 atm. R-1 and R-2 are two-stage Hoke regulators TThich maintain constant pressure on the upstream side of the flon-meters. Flow rater were determined by measuring the pressure drop across tubce filled with cotton. The pressure drop acros9 the cotton-plug restriction is almost a linear funcbion of the volume flow rate. Only at high flow rates does appreciable departure from linearity appear. The flow range can easily be changed. by removing or adding cotton. Calibration was effected by a displacement meter, The pressure drop across the rest,riction Tras indicated by a Meriam 20-inch...

Experimental measurements of the induction distances are given for hydrogen-oxygen, acetyleneoxygen, acetylene-air, methane-oxygen, carbon monoxide-oxygen, hydrogen-oxygen-nitrogen, hydrogen-oxygen-helium, hydrogen-oxygen-argon, and hydrogen-oxygen-carbon dioxide mixtures at initial pressures of 1, 5, 10 and 25 atm. An empirical relationship involving the combustion temperature, burning velocity, sonic velocity in the unburned gas and Reynolds number based on the burning velocity has been established to predict detonation induction distances of certain combustible gas mixtures. The correlation is not satisfactory for mixtures containing hydrocarbons; an explanation of this discrepancy is offered.

No abstract

No abstract

Composition, temperature, pressure and density behind a stable detonation wave and its propagation rate have been calculated for seven hydrogen-oxygen mixtures at 1, 5, 25 and 100 atm initial pressure, and at an initial temperature of 40 C. For stoichiometric mixtures the calculations also include an initial temperature of 200 C. According to these calculations the detonation velocities of hydrogen-oxygen mixtures increase with increasing initial pressure, but decrease slightly when the initial temperature is raised from 40 to 200 C. The calculated detonation velocities agree satisfactorily with values determined experimentally. These values will be published in the near future.T HE CALCULATIONS are based on the assumption that complete thermodynamic and chemical equilibrium is established in the wave. Dissipating effects such as viscosity, heat transfer by conduction and radiation, and chemical reaction rate phenomena are disregarded. The calculations were carried out for hydrogen-oxygen mixtures using a rigorous method developed by Edse (l). 3 Other gaseous reactants will be treated in a future paper.The detonation parameters are derived from the Hugoniot equation for the reacted gas mixture in equilibrium, and from the condition that the detonation velocity is the minimum wave velocity of the possible velocities for the given state of the combustible gas.^Thus the results are based on the Chapman-Jouguet point at which the velocity of the reacted gas relative to the detonation wave is equal to the equilibrium sonic speed (2) in the reacted gas behind the wave. These conditions can be calculated more readily than those for a frozen speed of sound for which the ratio of the specific heats k must be calculated (3). Method of CalculationAssuming that only the neutral species of H 2 0, H 2 , 0 2 , OH, 0 and H occur in the detonation wave of hydrogenoxygen mixtures, the Hugoniot equation can be written Pa Vt