This study investigates the influence of geometry parameters of dimples on the blade surface in a high-turning linear compressor cascade. Four arrays of dimples with a depth-to-diameter ratio of 0.25 and the depth varying from 0.1 mm to 0.6 mm in six different configurations were erected to explore their effects on loss coefficients, streamline patterns, topology structure as well as vortex structure. The dimple configurations were evenly distributed among 42–65% axial chord on the suction surface. The Reynolds-averaged Navier–Stokes method was used in the computational investigation under the nominal condition. The results show that the dimples with proper size can positively affect the flow loss and flow pattern. The larger dimples do not dramatically overtake the smaller ones in loss reduction, suggesting that there exists an optimal size of dimples in loss reduction within compressor cascades. The total number of singular points and separation lines in the cascade are both reduced under the dimple configurations. Furthermore, the vortex structures of the prototype and a prominent dimple configuration were depicted in detail. Passage vortex plays a dominant role in organizing the fluid within the cascade. The dimple configurations can suppress the migration of passage vortex along the spanwise direction and thereby decreases its influence region.

The flow characteristics and loss behavior over an array of parallel recessed dimples on a high turning linear compressor cascade have been investigated using the Reynolds-averaged Navier–Stokes approach. Steady simulations have been carried out at three dimple locations of 10–32%, 38–60%, 60–82% chord length of suction surface with the inlet Mach number of 0.7. Flow conditions were compared in exit loss coefficient, static pressure rise, streamline patterns, vortex structures, boundary layer parameters, and blade surface pressure between the smooth and the modified cascades. The results indicate that the dimples prior to the separation line report an overall enhancement in the aerodynamic performance in comparison to that of a smooth blade. Symmetric spanwise vortex, which energizes the boundary layer, can roll up inside the dimples. Therefore, the boundary layer with the higher momentum can bear the adverse pressure gradient, which will suppress the flow separation and associated losses. Three dimpled configurations can all eliminate the separation bubble on the suction side, but the dimples located at 60–82% chord length take the negative effect on the aerodynamic performance due to the more chaos condition in the corner separation region. The comparison results also indicate that the optimum location of dimples may exist in front of the separation bubble. Loss reduction of 18.8% and 10.8% can be achieved under the 10–32%c and 38–60%c dimple configurations, respectively.

Particle deposition often takes place in a gas turbine and it varies the surface roughness which has a great impact on flow and heat transfer characteristics. The accurate prediction of particle deposition and roughness caused by deposition is very important. In this study, one algorithm is proposed to calculate the roughness height resulting from the deposited particles which is predicted by the critical viscosity deposition model. The effect of roughness is coupled into governing turbulent equations through the specific dissipation by UDF. The particle deposition law, the corresponding roughness height, and its effects are discussed. The results show that the particles deposit along the leading edge and the deposition extends to the rear part after a strip of no-deposition region on the pressure side. On the suction side, there is much less deposition. The deposition on the tip is significantly affected by the tip flow characteristics including tip separation vortex and leakage vortex. The deposition is becoming more with the diameter of particle. The roughness height corresponding to the deposition is calculated and has similar pattern with deposition. Moreover, the number of particles deposited on the rough wall is more than that in other areas with time. This leads to more and more particles accumulated in the areas over time to form the roughness surface which then enhances the heat transfer. In return, the temperature influences the deposition and the roughness, all of which are undesirable in a turbine.

The influence of the ratio of dimple depth-to-print diameter ($\lambda$) on the highly loaded compressor cascade NACA0065-K48 is investigated based on the Reynolds-averaged Navier–Stokes (RANS) method. Simulations are conducted with a validated shear-stress transportation (SST) turbulence model coupled with the Gamma-Theta ($\mathsf{\gamma }-{\mathrm{Re}}_{\mathsf{\theta }}$) transition model at the inlet Mach number of 0.7. At 5~25% of the axial chord on the suction surface, four rows of dimples are arranged in parallel, and the dimples’ spacing is 4 mm. Moreover, there are five kinds of $\lambda$, ranging from 0.125 to 0.875, which determine the pressed arc of a spherical dimple. Three flow regimes (diffuser–confuser flow, tornado-like vortex and horseshoe vortex) with the same topological structure are observed in these dimples, which affect the flow and performance of the cascade by changing the energy distribution. The distribution of turbulent kinetic energy (TKE) reflects the disturbance of the tornado-like vortex in the inferior arc dimples ($\lambda =0.375$) intensely, whereas the disturbance of the horseshoe vortex in superior arc dimples ($\lambda =0.625, 0.875$) is relatively weak. Numerical results indicate that the loss of the corner separation can be reduced with a dimples array, which is mainly related to the vertical climbing of the lateral flow that delays the starting point of the corner separation and weakens the mixing process. However, the loss in the wake of the dimpled cascades increases, which is caused by the thickened boundary layer induced by the high turbulent vortices. The dimpled cascade with $\lambda =0.625$ can achieve the most significant loss reduction (13.47%), while ensuring the pressurization capacity.

In this study, to reduce the negative impacts of cavity leakage flow on the aerodynamic performance of a shrouded stator, the structure of the seal cavity is optimized. First, the cavity structure is parameterized, and under different cavity design schemes, annular cascades with a cavity structure are simulated numerically to construct a sample database. On this basis, a neural network and genetic algorithm are used to establish the relationship between the modeling parameters and performances to find an optimal solution. Finally, the mechanism whereby the optimized cavity improves the compressor performance is analyzed. The results show that the optimized seal cavity reduces the leakage flow by approximately 30%, while the windage heating through the cavity remains unchanged. Also, the circumferential velocity of the leakage flow at the outlet of the cavity increases, so its ability to resist the boundary layer and pressure gradient extrusion is enhanced in the cascade passage. These two improvements reduce the influence area of the cavity leakage vortex and delay the passage vortex, which can decrease the momentum loss thickness of the boundary layer and reduce the total pressure loss coefficient of the cascade by 2.14%. Furthermore, the changes in the optimized cavity structure exhibit a high degree of consistency with the trends of the correlation analysis. Therefore, according to the structural characteristics and control mechanism of the seal cavity, the proposed optimization method can obtain remarkably effective cavity structures, taking into account the cascade loss minimization as well as the security stability of the inter-stage sealing.