Examination of a large number of experiments dealing with flashing flows in converging and converging-diverging nozzles reveals that knowledge of the flashing inception point is the key to the prediction of critical flow rates. An extension of the static flashing inception correlation of Jones [16] and Alamgir and Lienhard [17] to flowing systems has allowed the determination of the location of flashing inception in nozzle flows with subcooled inlet conditions. It is shown that in all the experiments examined with subcooled inlet regardless of the degree of inlet subcooling, flashing inception invariably occurred very close to the throat. A correlation is given to predict flashing inception in both pipes and nozzles which matches all data available, but is lacking verification in intermediate nozzle geometries where turbulence may be important. A consequence of this behavior is that the critical mass flux may be correlated to the pressure difference between the nozzle inlet and flashing inception, through a single phase liquid discharge coefficient and an accurate prediction of the flashing inception pressure at the throat. Comparison with the available experiments indicate that the predicted mass fluxes are within 5 percent of the measurements.

A nonequilibrium vapor generation model for flashing flows is presented. The model consists of a flashing inception point, a bubbly flow regime followed by a bubbly-slug regime, an annular or annular-mist regime, and finally a dispersed-droplet regime. Existence of superheated liquid at the inception point and beyond is recognized. The vapor generation rate is calculated from the flow-regime dependent interfacial area density and net interfacial heat flux. However, the bubble number density at the flashing inception point was varied to obtain optimum fits with the void fraction data taken in a vertical converging-diverging nozzle. The interfacial area density at the inception point, thus determined, showed a rapid increase with the decrease in the liquid superheat at that point. This trend is plausible, since in the limit of thermal equilibrium flow where the liquid superheat approaches zero, the interfacial area for heat and mass transfer should be very large.

Three experiments done over twenty years on gaps in LX‐17 are described. For the detonation front moving parallel to the gaps, jets of gas products were seen coming from the gaps at velocities 2 to 3 times greater than the detonation velocity. A case can be made that the jet velocity increased with gap thickness but the data are scattered. For the detonation front moving transverse to the gap, time delays were seen. The delays roughly increase with gap width, going from 0–70 ns at “zero gap” to around 300 ns at 0.5–1 mm gap. Larger gaps of up to 6 mm width almost certainly stopped the detonation, but this was not proved. Real‐time resolution of the parallel jets and determination of the actual re‐detonation or failure in the transverse case needs to be achieved in future experiments.

No abstract