Strong Spherical Blast Waves in a Dust-Laden Gas

Abstract

separation angle (af) could be determined from the disappearance of the pressure overshoot downstream of reattachment on the wedge. This criterion is less reliable for the axisymmetric case because the attached flow pressure distribution is already of this form with a locally two-dimensional pressure rise at the corner followed by a gradual decay to the cone value far downstream on the flare. Incipient separation was determined here by three methods had given good agreement for the two-dimensional configuration.4'5 These were 1) looking for the onset of an increase in pressure and heat-transfer rate ahead of the flare; 2) looking for a change in the shape of the pressure and heattransfer distributions on the flare; and 3) by plotting the separation length (/) as measured from schlieren photographs, against flare angle and extrapolating to / = 0. The answers from these three methods agreed to within -|°. The results (Fig. 3) are compared with the wedge compression corner experiments of Elfstrom and Coleman.4'6 The results of Kuehn 2 for Mach numbers of 2 and 4 are also shown. All of the data exhibit the trend of decreasing a- with increasing Red where dL is the undisturbed boundary-layer thickness at the cylinder-flare junction. This is opposite to the trend noted by Rose, Page, and Childs in their nozzle wall experiments and may be the result of different upstream histories of the boundary layers in the two cases. Data are presented by Rose, Page, and Childs1 which indicate that separation occurs in axially symmetric flows at much lower pressure rises than would be expected on the basis of previous data for planar two-dimensional This conclusion is in direct contrast to the information given here and to that given by Kuehn, 2 both of compare planar and axisymmetric external flows. Either there is a great difference between incipient separation on internal and external surfaces, or the determination of incipient separation is very sensitive to the detection method employed, as found for example by Spaid and Frishett.7 From a practical viewpoint, the designer needs to know when significant changes in pressure and heat-transfer rate distributions will occur as a result of separation so there is considerable merit in using such measurements to define incipient separation whenever possible. To summarize, our results have shown that for axisymmetric external flow the incipient separation angle is slightly higher than for the planar two-dimensional case. There is a close similarity between the surface pressure and heat transfer rate distributions for the two cases.