A computer three-dimensional (3D) vision-aided performance prediction framework for turbomachinery is established in this paper, to improve the accuracy and generalization ability of the artificial neural network (ANN) model under inputs of more than 90 control parameters. In this framework, a RandLA-encoder is built to extract the flow information related to performance and geometric parameters from point cloud data of flow fields inside impellers. By implicitly learning this kind of flow information, the prediction error of the ANN model is reduced by 20–30% compared with the traditional one. Based on this, a full-3D optimization with 91 variables, including arbitrary blade surface and non-axisymmetric (but periodic) hub surface, is conducted on Krain low-speed impeller, aiming at a comprehensive performance improvement. After the optimization, compared to the baseline, the maximum isentropic efficiency of the compressor is increased by 1.6%, the isentropic efficiency at design point is increased by 1%, and the flow range is increased by 5%, with a slight increase in pressure ratio.
A full-range prediction model for turbomachinery based on the flow-field information code is established in this article to solve the problems that traditional models do not have enough prediction accuracy and cannot reflect the complete performance characteristics of the impeller. The model, which can predict the complete performance curve of the impeller with higher accuracy, consists of two multilayer artificial neural network (ANN) submodels. Different from the traditional model, the ANN submodel uses the flow-field information code for pretraining layer by layer. The flow-field information code is the characteristic information extracted from the impeller flow field through the proper orthogonal decomposition (POD) method. By implicitly learning the flow-field information, the prediction error of the model is reduced by 29.7% compared with the single hidden layer ANN. Based on this model, the nonaxisymmetric, but periodic, hub optimization of a centrifugal impeller with 30 variables is carried out, with the goals of the higher efficiency and the wider flow range at the specified pressure ratio and the massflow rate at the design point. The result shows that, after the optimization, the isentropic efficiency at the design point increases by 1% and the flow range increases by 2% compared to the baseline.
Applying an appropriate roughness model is one of the most effective ways to improve the accuracy of the numerical simulation for centrifugal compressors, and it is also the key to the study of surface deterioration such as compressor fouling and corrosion. The basis for using the roughness model is to determine the conversion relation between physical roughness and equivalent roughness. In this paper, numerical and experimental studies on the roughness conversion relation for the centrifugal compressors were performed by means of a unique self-developed “pseudo front stage (PFS)” device. An equivalent roughness conversion coefficient of 8.5 was determined on the basis of fitting performance of the U-bend and return channel model. The validity of the conversion coefficient in improving the performance prediction accuracy of the centrifugal compressors was verified by three centrifugal compressor models. The concept of weighted average roughness was proposed to quantify the roughness sensitivity and roughness level of the centrifugal compressor. Quantitative research on the roughness effect of centrifugal compressors indicated that the peak efficiency degradation due to the roughness effect in the compressors was linearly associated with the logarithm of the relative roughness when the impeller of the centrifugal compressor was in the hydraulically rough regime. Doubling or halving the roughness value of all components of a centrifugal compressor in the hydraulically rough flow regime will result in a change in the peak efficiency of the compressor by about 1.72%.
The development of free-stream disturbances in flow over a vertically vibrating flat plate with a slender leading edge is investigated. The evolution of the optimal inflow perturbation that results in the maximum amplification is computed to investigate the effect of the plate vibration on the development of free-stream disturbance, secondary instability of streaks and subsequently the bypass transition to turbulence. It is observed that the plate vibration leads to periodic change of the angle of attack, shifting the free-stream disturbance to the upper or lower side of the plate. Therefore, the development of steady inflow perturbations, which receive the largest amplification, is interrupted by the vibration, and the perturbation amplification via the lift-up mechanism is weakened. The vibration brings a second peak of perturbation growth at the vibration frequency, leading to high-frequency free-stream perturbations penetrating into the base boundary layer, which is not observed in flow over a stationary plate owing to the sheltering mechanism. This resonance of the flow perturbation and the vibrating plate is explained by the staggering effect of the leading edge. Further, the direct numerical simulations with the optimal inflow perturbation imposed on the inflow boundary show that the vertical vibration of the plate leads to streamwise periodic vorticity near the edge of the boundary layer. This inhomogeneity of the streamwise vorticity brings about streamwisely localized distortion of the low-speed streaks and, thus, an intermittent secondary instability. Therefore, before the streaks break down to turbulence, they undergo several rounds of secondary instabilities, resulting in an elongated bypass transition process.
In recent years, in order to achieve higher performance, the number of design variables used in the aerodynamic optimization of turbomachinery (based on surrogate models) has increased significantly. However, the increase in the design dimensions brings the problems of sparse distribution of training samples and numerous inaccurate local optima to the surrogate model, making it difficult to obtain an accurate Pareto front in high-dimensional aerodynamic optimization. To solve this, adaptive batch sampling strategy (ABSS) is proposed in this paper. By searching numerous local optimum points with large prediction errors in the surrogate model, ABSS can provide batches of valuable samples for each iterative update of the model. Compared with the current model update strategy based on the best point, which requires hundreds or thousands of iterative model updates, ABSS only needs to go through a few model updates to make the predicted Pareto front close to the real one. ABSS not only greatly shortens the whole optimization time, but also makes it easier to jump out of the local optima for the optimization. Based on this, a 114-dimensional aerodynamic optimization of the full three-dimensional centrifugal impeller (including arbitrary blade surfaces, independent splitter surfaces, and non-axisymmetric hub surfaces) is carried out. The results show that, compared with the baseline impeller, the design point isentropic stage efficiency of the optimal impeller is increased by 2.1%, the design point pressure ratio is well controlled to 1.84, and the choke margin is increased by 9.2%. This research breaks through the largest number of variables in the current centrifugal impeller optimization, proposes the optimization methods of the full-3D centrifugal impeller, and also provides a reference for the future high-dimensional aerodynamic optimization of turbomachinery.
11 March 2024 in Journal of Engineering for Gas Turbines and Power
Aerodynamic instability plays an important role in compressor design and may cause performance degradation and fatigue damage. In this paper, an experimental study on the evolution of aerodynamic instability is carried out on a compressor that combines the performance benefits of an axial stage and centrifugal stage. The spatiotemporal characteristics of unsteady wall pressure were obtained using fast-responding pressure transducers over a range of operating conditions. The results show that the axial stage works on the positive slope of the performance characteristic curve from choke to stall at low-speed operating conditions, and mainly features rotating instability. Rotating stall is also observed in the impeller (IMP) and diffuser passages. At medium-speed operating conditions, the centrifugal stage suffers a high-frequency mild surge, alternating with rotating stall. With the increase in back pressure, the mild surge diminishes, and rotating stall persists. This behavior is similar to a two-regime-surge, which has been reported for centrifugal compressors. At high-speed operating conditions, the compressor directly reaches surge without other instabilities. Further analysis of the spatial pattern of the rotating stall revealed the existence of a high-pressure region near the volute tongue, resulting in obvious pressure distortion along the circumferential direction at the volute inlet. This induced the amplitude difference of stall cells in corresponding diffuser passages. The disturbance caused by stall cells propagates upstream through the blade passage, and the largest pressure disturbance induced by the stall cell propagation appears in a circumferential position 45 deg downstream of the volute tongue at the impeller inlet and the axial stage inlet.