Research Of Evvard Method About Two-dimensional Supersonic Nozzles With Continuous Wall Curvature

With the rapid development of science and technology, advanced aircraft has characteristics of high-speed, accuracy, long-distance and invisibility. High-speed is the most remarkable feature, which demands the ground experiment facilities for high-speed aircraft test like supersonic tunnels with high performance and flow field quality. As the major component of a supersonic wind tunnel, nozzle plays an important role in the forming of supersonic flow. The quality of nozzle design directly decides the test section quality of the wind tunnel. Domestic method of the nozzle design is mainly based on Sivells method, which can fulfill the needs of most wind tunnel tests, but it is not quite enough for high quality supersonic tunnels with refined flow field. While comparing to wind tunnels with high quality and strict standards in foreign countries, there still have great potential to improve supersonic nozzle design.Nozzle design method tends to be solidified in the 60 s of the last century and there has been no big innovations appear since then. Before Sivells method, Evvard proposed a nozzle design method, where he introduced the main aspects of the method in principle, showing it can achieve a better flow field. In recent years, it shows that Evvard method have some advantages in flow field uniformity, as axial Mach number distribution. In order to promote the quality of wind tunnel flow field and to have much more accurate ground experiment data of high speed aircraft, research on Evvard method is necessary and urgent.Based on existing methods, with the transonic solution of throat, and given axial Mach number distribution, nozzle profile, which is mentioned in Evvard method, can be calculated. Firstly, Method of Characteristics(MOC), which is the most fundamental principle, is implemented as a calculation program. Secondly, Evvard method is implemented, which is different from others’ methods, abandons the radial flow assumption around throat, and simulates the real flow situation directly. Usually the solution of throat has many ways just like Sauer method, Hall method and Kliegel method. They have different application conditions. Here Sauer transonic solutions of throat are chose for two reasons, easiness of calculation and that the errors can be controlled. Thirdly, the axial Mach number distribution according to the solution of throat and the first rhombus zone is chose by numerical simulation. Professor Yi offers 6rd order Bezier curve to simulate the axial Mach number distribution. The numerical simulation results show that the differences in the axial Mach number gradient of the first rhombus zone and the root square deviation of Mach number(RSM) of test section profile can be neglected. But the differences in the wall Mach number distribution and axial Mach number distribution are obvious. The optimized of 6rd order Bezier curves are used in the later research. Fourthly, the boundary conditions of the first rhombus zone, which can be measured by the size of test section and the design Mach number uniquely, are selected. Lastly, the nozzle profiles are obtained according to the MOC with the three boundary conditions and compared with results from Sivells method.The flow field of nozzle profile obtained by this paper is compared with the profile obtained by Sivells method. The results show that several indexes, such as axial Mach number gradient in first rhombus zone, RSM of nozzle outlet profile, RSM of test section central profile and RSM of test section outlet profile, obtained by the new method are better than Sivells method. Although the new method is not verified with experimental results, it could play an instructive role and lay the foundation for the future works.