Hydrodynamic design of contraction, diffuser and elbow geometries for a high speed cavitation tunnel

In the literature, hydrodynamic design of high speed and closed circuit cavitation tunnel in which hydroacustic, hydrodynamic and cavitation performance are investigated with controlled pressure is an essential research study. Especially, hydrodynamic design of contraction, diffuser and vaned elbows are main parts of the cavitation tunnel that control flow quality inside the test section, and also they directly affect main dimensions of the cavitation tunnel. In this thesis, hydrodynamic design of contraction, diffuser and vaned elbows was done to gain reference knowledge for the cavitation tunnel, which will be brought in our faculty by KATMANSIS (Cavitation Tunnel and Manoeuver Experiment System) project. The tunnel efficiency is determined with no flow separation and minimum energy loss for each part of the tunnel. Particularly, control of minimum reachable cavitation number, turbulence intensity and flow uniformity at all directions outside of the boundary layer inside the test section are the basic design criterions during design stage. The literature investigation about cavitation tunnels and its sections, which were investigated in the thesis, are presented in the first chapter. In the second chapter, general design criterions are given with desired limits in the test section, for instance minimum cavitation number should be around 0.2, flow uniformity should be less than 1% at all directions with respect to the reference flow speed, turbulence intensity should not exceed 1% and minimum boundary layer development should be provided. The contraction geometry, which directly controls flow quality in the test section and takes place just before test section, was investigated with four different parameters. These are contraction ratio, contraction length and contraction wall profile which controlled second derivative of contraction entrance and the position of inflection point. Diffuser was controlled with the expand ratio and diffuser angle to be able to discharge water from the test section smoothly. This section does not only affect length of the tunnel, but also determines the average flow speed in parts of downstream side of diffuser with its expand ratio. Hence, diffuser should be properly designed to minimize total pressure loss of the tunnel. Finally, the vaned elbows, which take place at downstream of diffuser and upstream side of contraction, were designed with profile shape, profile length, attack angle and profile number ratio. These criterions' effects on cavitation tunnel are referred. Design studies were done via using Computational Fluid Dynamics method. In third chapter, theoretical background on CFD was given with its historical progress. Furthermore, the details about finite volume method take place in the same chapter, SIMPLE method for velocity-pressure discirizastion and SST k-ω that chosen as turbulence model. All studied cases' results for the section, which was mentioned in the second chapter, was presented in fourth chapter. Flow speed in test section was set as 10m/s, which is determined as service speed in test section, for all cases at design stage. However, for the cases considered as optimal geometry also analyzed for other two different flow speed, which are 2m/s for low Reynold's Number and 15m/s for maximum desired velocity in test section. First cases were done by studing on contraction ratio which was taken as 5 and 6. Then their effect in test section was indicated. Positive effect of increasing contraction ratio was showed with flow uniformity and turbulence intensity as expected while boundary layer flow separation were occurred around contraction entrance. Thus, contraction ratio 5 was chosen as optimum contraction ratio. Incident pressure change caused adverse pressure gradient with short contraction length, so boundary layer flow separation was generated in the case whose ratio of contraction length to test section height was L_c/h=4.6. Thus, the case with longer contraction length with L_c/h=4.75 gave also good result with no separation. However, there were no longer contraction length since it is not acceptable having thicker boundary layer in test section. It was preferred to derive contraction wall profile from 6th order four-term polynomial. Hereby it is possible to control second derivative of contraction entrance profile as well as inflection point position on contraction. The position of inflection point directly affect minimum value of pressure coefficient in test section, while second derivative of wall profile affects flow quality around contraction entrance. When inflection point gets close to test section, reachable minimum cavitation number was increasing in test section, on the other hand whet it gets close to contraction entrance, boundary layer flow separation started to occur. Thus, it was decided to locate inflection point to (x_c^i)⁄L_c =0.5 with no flow separation and reachable minimum cavitation number. In addition, three different values of second derivative of contraction entrance were analyzed, only d_2=0 was acceptable due to having no flow separation contrary to other values. In chapter 4.2, study on diffuser was examined by controlling two main parameters which are expansion ratio and diffuser angle. Optimum diffuser geometry was designed with minimum pressure loss and diffuser length by controlling the parameters directly bounded with each other. It was observed that geometries with fixed diffuser angle did not give expected result during design stage. Hence, chamfer was defined to corner of sections along the tunnel and it improved the results. According to diffuser length, the results were not enough, so the study continued with stepped diffuser geometry, which was divided into three equal parts along the diffuser, each part of diffuser was controlled with its own individual diffuser angle. Finally, optimum diffuser geometry was designed which has expansion ratio 2.8 and diffuser length 19.5m. Finally, 2D vaned elbow geometries' results were given in Chapter 4.3. Profiles' geometry substantially affects flow in this zone. NACA9410, which was thought as the most probable geometry to direct the passing flow with its shape among all NACA four digit profiles, was analyzed and large flow separation detected at downstream side of profile. Thus, two original profile geometry designed for the tunnel. Furthermore, studies about vaned elbow contained three more parameters for elbow design. These are profile number ratio, profile length and profile attack angle, in order to have minimum pressure losses and provide maximum flow uniformity. Eventually wet surface area is increasing related with increasing profile number and profile length. Different cases were investigated to have minimum pressure loss. Vaned elbows geometry before contraction and after diffuser were design by getting experience in this chapter. So, final values of parameters of vaned elbow geometry are Prototipe-2 profile, 2.2 profile number ratio, which is hypotenuse length to profile number in elbow, profile length L_P=0.75 and attack angle α=5 The result of final case that contains all designed geometry of sections was analyzed in one flow volume for three main speed in Chapter 4.4. Thus, flow in test section was modelled more accurately and flow qualites were reported with including all design geometries' interaction with each other. It is showed that all desired values of criterion can be provided by putting honey cumb geometry before contraction. Moreover, maximum power requirement of the tunnel was calculated according to pressure loss of 15m/s case and empirical calculation for the parts that was not included to flow volume such as lower parts of the tunnel.