Supersonic diffuser

Fig. D-2 shows the shock-wave formation at a supersonic diffuser composed of a divergent nozzle. Three types of shock wave are formed at three different back-pressures downstream of the diffuser. When the back-pressure is higher than the design pressure, a normal shock wave is set up in front of the divergent nozzle and the flow velocity becomes a subsonic flow, as shown in Fig. D-2 (a). Since the streamline bends outwards downstream of the shock wave, some air is spilled over from the air-intake. The cross-sectional area upstream of the duct becomes smaller than the cross-sectional area of the air-intake, and so the efficiency of the diffuser is reduced. The subsonic flow velocity is further reduced and the pressure is increased in the divergent part of the diffuser.

Figure D-1. Pressure and flow velocity in a subsonic diffuser and a supersonic diffuser.

Figure D-2. Normal shock-wave formation of a supersonic diffuser at different back-pressures (a) high back-pressure, (b) optimum backpressure, and (c) low back-pressure.

An internal compression system forms several oblique shock waves and one normal shock wave inside the duct of the air-intake. The first oblique shock wave is formed at the lip of the air-intake and the following oblique shock waves are formed further downstream the normal shock wave renders the flow velocity subsonic, as shown in the case of the supersonic diffuser in Fig. D-1. The pressure recovery factor and the changes in Mach number, pressure ratio, and temperature ratio are the same as in the case of the external compression system. Either external or internal air-intake systems are chosen for use in ramjets and... 

Continuous-wave HF/DF lasers are designed with elements best illustrated by considering a simple combustion-driven supersonic diffusion laser as an example. In this case, no external power supply is required the laser operates purely on chemical energy. Figure 5 presents the main elements that make up this device. The laser combustor... 

The ejector is widely used as a vacuum pump, where it is staged when required to achieve deeper vacuum levels. If the motive fluid pressure is sufficiently high, the ejector can compress gas to a slightly positive pressure. Ejectors are used both as subsonic and supersonic devices. The design must incorporate the appropriate nozzle and diffuser compatible with the gas velocity. The ejector is one of the ( to liquid carryover in the suction gas. 

We have so far assumed that the atoms deposited from the vapor phase or from dilute solution strike randomly and balHstically on the crystal surface. However, the material to be crystallized would normally be transported through another medium. Even if this is achieved by hydrodynamic convection, it must nevertheless overcome the last displacement for incorporation by a random diffusion process. Therefore, diffusion of material (as well as of heat) is the most important transport mechanism during crystal growth. An exception, to some extent, is molecular beam epitaxy (MBE) (see [3,12-14] and [15-19]) where the atoms may arrive non-thermalized at supersonic speeds on the crystal surface. But again, after their deposition, surface diffusion then comes into play. 

Despite the fact that relaxation of rotational energy in nitrogen has already been experimentally studied for nearly 30 years, a reliable value of the cross-section is still not well established. Experiments on absorption of ultrasonic sound give different values in the interval 7.7-12.2 A2 [242], As we have seen already, data obtained in supersonic jets are smaller by a factor two but should be rather carefully compared with bulk data as the velocity distribution in a jet differs from the Maxwellian one. In the contrast, the NMR estimation of a3 = 30 A2 in [81] brought the authors to the conclusion that o E = 40 A in the frame of classical /-diffusion. As the latter is purely nonadiabatic it is natural that the authors of [237] obtained a somewhat lower value by taking into account adiabaticity of collisions by non-zero parameter b in the fitting law. 

In a diffusion pump, the dense oil vapour produced by the boiler (see Fig. 1.12) is ejected into the vacuum at high (or even supersonic) speed through the nozzles. 

When the tube is closed at one end and ignited there, the propagating wave undergoes a transition from subsonic to supersonic speeds. The supersonic wave is called a detonation. In a detonation heat conduction and radical diffusion do not control the velocity rather, the shock wave structure of the developed supersonic wave raises the temperature and pressure substantially to cause explosive reaction and the energy release that sustains the wave propagation. 

In his supplement to Syllabus, Dunkle remarked (Ref 22, p lid) that schlieren photography of deton waves in 40/60 C2H2/O2 initially at 1/4 atm showed a wavy pattern of criss-crossing dark diffuse lines behind the front. Fay Opel (Ref 17) calculated that if these lines are a weak wake of Mach waves in supersonic flow, the flow of the burnt gases with respect to the front is Mach 1.14 rather than Mach 1.00 as in a C-J process. However, at this pressure the reaction is complete within a fraction of a millimeter behind the front, and the flow could very well accelerate to Mach 1.14 with density decrease below the C-J value. Fay Opel traced the effect to the boundary layer. 

Notations 1 - swirling device, 2 - sub/ supersonic nozzle, 3 - working section, 4 -device for gas-liquid mixture withdrawal, 5 - diffuser... 

Operation. In a diffusion pump, the pump fluid is heated so that a vapour pressure of 1-10 mbar is established in the boiler. The vapour rises in the jet assembly where it is expanded through nozzles and enters the space between the nozzle and the cooled wall of the pump at high supersonic velocity. Pumping action is based on the transfer of momentum in collisions between the high speed (several times the speed of sound) pump fluid vapour molecules and particles that have entered the vapour jet. 

The basic principle of a diffusion pump can be explained with a simple single-stage mercury diffusion pump (see Fig. 7.21). On the system side of the pump (at about 10 2 to 10 3 torr, or better), gas molecules wander around, limited by their mean free path and collisions with other molecules. The lowest section of this diffusion pump is an electric heater that brings the diffusion pump liquid up to its vapor pressure temperature. The vapors of the diffusion pump liquid are vented up a central chimney where, at the top, they are expelled out of vapor jets at supersonic speeds (up to 1000 ft/sec). Below these jets is a constant rain of the pumping fluid (mercury or low vapor-pressure oil) on the gases within the vacuum system. Using momentum transfer/ gas molecules are physically knocked to the bottom of the pump, where they are trapped by the vapor jets from above. Finally, they are collected in a sufficient quantity to be drawn out by the auxiliary (mechanical) pump. 

Momentum transfer is possible because momentum = (mass) x (velocity), and the velocity of the diffusion pump fluid (when it leaves the pump jets) is typically at supersonic speeds. Thus, it is easy for the diffusion pump liquid to knock around any size molecule. 

In a mercury diffusion pump, the mercury is heated to the point of vaporization. This vapor travels up into the condenser area where it is ejected at supersonic speeds from little holes. The vapor knocks any wandering gas molecules down toward the mechanical pump outlet which can then expel them from the system. The vapor later condenses and collects in the heating pot for reuse. 

---