No abstract

It is essential for engineering manufacture and robust design to evaluate the impact of manufacturing variability on the aerodynamics of compressor blades efficiently and accurately. In the paper, a novel quadratic curve approximation method based on the scanning points of blade design profiles was introduced and combined with Karhunen–Loève expansion, a mathematical dimensionality reduction method for modeling manufacturing variability as truncated Normal process was proposed. Subsequently, Sparse Approximation of Moment-based Arbitrary Polynomial Chaos (SAMBA PC) and computational fluid dynamics (CFD) were applied to build a computational framework for stochastic aerodynamic analysis considering manufacturing variability. Finally, the framework was adopted to evaluate the aerodynamic variations of a high subsonic compressor cascade under the design incidence. The results illustrate that the SAMBA PC method is more efficient than the traditional methods such as Monte Carlo simulation (MCS) for stochastic aerodynamic analysis. Through uncertainty quantification, the impact of manufacturing variability on the global aerodynamic performance is primarily reflected in the fluctuation of aerodynamic losses, and the fluctuation of the total losses is mainly contributed by the fluctuation of the separation loss after the suction peak (a negative pressure spike near the leading edge (LE)) and the boundary-layer loss on the suction surface (SS). With sensitivity analysis, the most important geometric modes to aerodynamics can be revealed, which provides a useful reference for manufacturing inspection process and helps reduce computational cost in robust design.

In view of the characteristics of flow separation in the compressor cascade corner region, a new flow control method for installing little blades in the front of the cascade passage was proposed, which took into account the flow control advantages of end wall fences and vortex generators. Firstly, the little blades could hinder the cross flow on the end wall and the development of the horseshoe vortex pressure surface branch. Secondly, the little blades could generate induced vortices to take away the low-energy fluid near the end wall and the corner region. Based on numerical simulations, the effects of different pitchwise positions, stagger angles and heights of the little blades on the aerodynamic performance of the cascade were studied, and the optimal little blades were obtained by NSGA-II using EBF neural network as the agent model. The results show that the little blades have the optimal pitchwise position, stagger angle and height range for improving the aerodynamic performance of the cascade. When the optimized little blades are introduced in the baseline cascade, the stable working range of the cascade is expanded, and the stall type of the cascade changes from the hub-corner stall to the overload of flow separation near the mid-span. At the near stall attack angle of the baseline, the total pressure loss coefficient is reduced by about 10.38% and the static pressure coefficient is increased by about 4.31%. Meanwhile, the loss of the lower span is decreased and the diffuser capacity of the whole span is improved. The passage secondary loss and wake loss are reduced because of the delay of corner separation. Moreover, the strength of the end wall vortex is weakened and the end wall vortex no longer develops as part of the passage vortex. The induced vortex, horseshoe vortex pressure surface branch and initial passage vortex develop into new passage vortex.

Flow variations at the inlet boundary due to the compressor operational condition changes and geometric variations of the realistic compressor blades due to the manufacturing variability cannot be absolutely avoided, the global and local performance impact of which requires to be considered in the mechanism study of performance change and the aerodynamic shape design. In this paper, a method to analyze the simultaneous impact of the inflow Mach number, inlet incidence and geometric uncertainties was proposed. To make the uncertainty modeling of geometric variations faster and closer to engineering practice, a parametric mathematical model based on scanning points on the blade was introduced to describing the profile and torsion errors in the method specially. Meanwhile, a sparse grid based-Non-Intrusive Polynomial Chaos (NIPC) was used for uncertainty quantification and uncertainty sensitivity analysis to alleviate the computational burden. Then, the method was combined with a loss source calculation method to estimate the global and local aerodynamic loss changes of a controlled diffusion compressor blade in a reference flow state of high inflow Mach number and large positive incidence, and the response performance of sparse grid-based NIPC was verified. The results show that inlet incidence and torsion error have a significance uncertainty effect on the boundary-layer separation above the suction surface, which is main reason for the fluctuation of global aerodynamic loss. The uncertainty effect of profile error on the boundary-layer separation is relatively weak, but profile error could have a certain uncertainty effect on the leading edge separation. The boundary-layer separation is insensitive to Inflow Mach number. Furthermore, a stochastic aerodynamic analysis in different reference inflow states was investigated, which reveals some laws that the uncertainty of the aerodynamic losses and the sensitivity of the inflow and geometric uncertainties change with reference inflow states.

The axial location of full-span boundary layer suction is studied to explore the influences of suction slot on the cascade performance. At the design condition, the slot with 50% axial location shows a superior capability to reduce the total pressure loss. At the near stall condition, the more upstream of the suction slot is moved, the more total pressure loss is reduced, and the suction slot with a location of 0.7 axial chord length cannot effectively reduces the total pressure loss in all conditions. Moreover, a rearranged segmented suction slot according to the distribution characteristics of the flow reversal region is developed and compared with full-span boundary layer suction. The segmented suction slot shows significant advantages in delaying the stall occurrence, and the stall point is delayed from 7.9° to 10.0° compared with the baseline. According to a quantitative analysis method selected to measure the performances of flow control technologies, the wake loss is significantly reduced by the segmented suction slot. Finally, a set of micro-vortex generator is introduced in the cascade with a segmented suction slot, and the conclusion indicates that the portion near the end-wall is very effective to reduce the flow loss.