This research investigates the process of starting and breakdown of second throat diffuser during high-altitude test of conical nozzle with a high expansion ratio. The subscale experimental setup includes a conical nozzle with an expansion ratio of 53, plus a second throat diffuser with a contraction ratio of 1.85, using compressed air as the working fluid. Numerical simulation has been employed to identify the flow physics during the unsteady process of diffuser startup and breakdown. According to the results, the starting process is divided into three distinct stages. In the first stage, the exhausted flow from the nozzle enters the vacuum chamber, leading to an increase in vacuum chamber pressure. The second stage, corresponding to the period before full flow establishment in the conical nozzle, exhibits a relatively constant slope. In the final stage, associated with the transition from over-expanded to under-expanded states, the slope of the rate of evacuation development descends. Next, the vacuum degradation in termination process has been analyzed, and it has been found to include three stages: high slope, middle slope, and low slope. The flow physics during the start process, similar to the results observed in other conical nozzles, exhibits only a Mach reflection structure. However, during the termination process, the flow physics involve a combination of both structures, including Mach reflection and Cap Shock. The results indicate that during the start, an internal shock only interacts with the separation shock, and no special change occurs in the Mach reflection structure. In contrast, during the termination process, unlike what has been reported in previous studies on conical nozzles, the structure of cap shock waves and restricted shock separation patterns are also observed. Another distinction between the starting and termination processes is related to the pressure distribution in the diffuser wall. The wall pressure at the diffuser inlet during the termination process has been reported to be 90% higher than during the starting process.
Read full abstract