• 대한기계학회논문집B
• /
• 제28권10호
• /
• pp.1231-1237
• /
• 2004

Determination of geometric design parameters of a second-throat type annual supersonic ejector is described. Tested geometric parameters were primary nozzle area ratio, cross-sectional area of second-throat, L/D ratio of second-throat and primary flow injection angle. Varying these four geometric parameters, we build a test matrix made of 81 test conditions, and experimental apparatus was fabricated to accommodate them. For each test condition, the stagnation pressure of primary flow and the static pressure of the secondary flow were measured simultaneously along with their transition to steady operation and finally to unstarting condition. Comparing the performance curve of every case focused on starting pressure, the unstarting pressure and the minimum secondary pressure, we could derive correlations that the parameters have on the performance of the ejector and presented the optimal design method of the ejector. Additional experiments were carried out to find effects of temperature and mass flow rate of the secondary flow.

• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2005년도 제25회 추계학술대회논문집
• /
• pp.46-55
• /
• 2005

For the sea-level performance test of rocket motor designed to operate in the upper atmosphere, ejectors with no induced secondary flow are generally used, which serves dual purposes of evacuating the test cell and performing as a supersonic exhaust diffuser (SED). The main concern of this research is to simulate starting transients in order to visualize evolution of internal shock structures in SED with different initial cell (vacuum chamber) pressures. RANS code with low Reynolds $k-\varepsilon$ turbulence model was employed for these computations. Numerical results were compared with the pressure measurements previously performed [Proceedings of 2004 Annual Conference, KIMST], and showed good agreements with pressure-time history of measured data. In the case of low vacuum chamber pressure, abrupt impingement of the under-expanded supersonic jet from the nozzle onto the diffuser wall was observed, whereas initial impingement point was located downstream and moved slowly upstream in the case of non-vacuum chamber pressure. In spite of initially dissimilar evolution of shock structures, iso-mach contour revealed that the steady shock structures had little difference except the location of flow separation and normal shock.