Investigation of non-axisymmetric endwall contouring in a compressor cascade

Liu, Xiwu; Jin, Donghai; Gui, Xingmin

doi:10.1007/s11630-017-0966-z

Cited by 12 publications

(5 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As people's living standards are improved, the requirements for using electrical devices and other performance are becoming higher and higher. However, the failure caused by compressor defects cannot be ignored [1,2]. So far, many studies have been done regarding the defect detection and fault diagnosis of compressor.…”

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Networks Combined with Machine Vision for Mechanical Compressor Defect Detection

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The study aims to detect the defects in the production line of compressor, promote the development of convolution neural network (CNN) in defect diagnosis and recognition, and expand the application of intelligent algorithm tools in the detection and recognition of defect and fault. The detection and recognition of the defects in the compressor workpiece were discussed based on the optimization of CNN. First, the active learning under the background of machine learning (ML) was introduced into the selection of sample training model for the annotation of massive data sets in diagnosing the compressor defect, and compared with the random selection method. Second, aiming at the limitation of deep CNN, through the introduction of deep separable convolution and reverse residual structure, an improved space-separable residual CNN model was proposed, and its training process was observed and analyzed. Finally, the improved space-separable residual CNN model was applied to the defect detection and recognition of compressor workpiece, and the effect of defect recognition was evaluated. The results showed that the accuracy of the active learning was higher than that of the random selection, and it can save about 18.76% of the cost of manual annotation data. The recognition accuracy of the improved CNN model is more than 90%, and the accuracy curve of the training set and the test set basically coincided in the later training period. The average recognition accuracy of the compressor defect detection model was 96.87%, and the normal workpiece could be fully recognized. The combination of improved and optimized CNN and ML has great potential in the detection of compressor defect.

show abstract

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Networks Combined with Machine Vision for Mechanical Compressor Defect Detection

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…Reising and Schiffer 12,13 found the direction change of the pressure gradient turning less vertical to the endwall streamline contributed to the performance improvement. Liu 14 and Li 15 found that the groove near the leading edge mostly influenced the adverse pressure gradient and improved the compressor performance. Cao 16 proposed that the CEW introduced an increased spanwise inward pressure gradient, which was beneficial to the performance.…”

Section: Introductionmentioning

confidence: 99%

Influence of the inlet boundary layer on non-axisymmetric endwall contouring effects in a linear compressor cascade

Teng

Zhu

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

The modern aero-engine compressors are commonly designed to be highly loaded which can introduce strong three-dimensional flow separation, and effective flow control is of vital importance in improving compressor performance. Non-axisymmetric endwall contouring nowadays is conducted on axial compressors to control the flow separation. In this paper, contoured endwall is designed based on a linear CDA compressor cascade. The Genetic Algorithm and ANN surrogate model are used in the optimization process of the hub contouring, based on RANS simulations. Three inlet boundary layers which follow the fully turbulent boundary layer assumption are generated with different thickness and the influence mechanism are numerically investigated. The optimal endwall with streamwise concave near suction side provides great performance improvement at all inlet boundary layers at the aerodynamic design point. The endwall flow near the suction side is accelerated by the contoured endwall with cross-flow pressure gradient altered, suppressing the corner separation. The main pressure loss is closely related to one streamwise vortex that blocks the endwall flow. Thicker inlet boundary layer can induce early forming and increased strength of the streamwise vortex, enhancing the cross-flow movement and its interaction with the blade surface flow. In addition, the pressure loss near endwall in the inlet boundary layer is found more dominant on endwall contouring effect than the boundary layer thickness.

show abstract

“…Chu 21 proposed an endwall design theory based on the velocity modification and achieved 1.65% reduction in numerical total pressure loss on a highly loaded linear compressor. Liu 22 designed CEW in a low-speed linear compressor cascade and found that the adverse pressure gradient was mostly influenced by the groove structure near the blade leading edge.…”

Section: Introductionmentioning

confidence: 99%

“…Based on the compressor configuration in Liu, 22 Liu 23 and Guo 24 studied the influence of solidity and camber on the design of endwall contouring, and concluded that the endwall contouring reduced the total pressure loss with different solidities, and the control effect was stronger when the camber increased. Wang 25 studied endwall contouring and tested it in a large size research compressor with embedded stators and showed that the strategy to minimize the pressure above 90% span was best to decrease stator loss.…”

Section: Introductionmentioning

confidence: 99%

Effective strategy of non-axisymmetric endwall contouring in a linear compressor cascade

Teng

Zhu

et al. 2021

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

The corner separation and the related secondary flow have great impact on the compressor performance, and non-axisymmetric endwall contouring is proved effective in improving compressor efficiency. The aim of the study is to improve the compressor performance by two local endwall contouring strategies at the design and off-design conditions. The endwall is parameterized and the Bezier curve is used to loft the endwall surface. The design of the contoured endwall is based on a multi-point optimization method to minimize the aerodynamic pressure loss. In order to identify the influence of the contoured endwall, a detailed flow analysis is conducted on four effective contoured endwall designs. The selected endwall geometries exhibit great control ability on the corner separation and significantly reduce the pressure loss at the two operating conditions. The directional concave near the leading edge can induce strong streamwise pressure gradient and accelerate the endwall flow, greatly reducing the cross-passage pressure gradient. The convex structures near the concave edge and at the outlet can block the cross-flow and prevent the interaction between the cross-flow and the suction corner flow. The benefit of the contoured endwall is mainly due to the re-distributed endwall static pressure and blocking of the cross-flow movement. In terms of the control effect, the shape of the concave also matters and better control effect is observed on the deep and wide concave. The flow will be guided by the concave, and the best suppression on corner separation is observed on the concave which follows the suction side. The results also indicate that the relief of the hub corner separation slightly increases the shroud pressure loss.

show abstract

Investigation of non-axisymmetric endwall contouring in a compressor cascade

Cited by 12 publications

References 5 publications

Convolutional Neural Networks Combined with Machine Vision for Mechanical Compressor Defect Detection

Convolutional Neural Networks Combined with Machine Vision for Mechanical Compressor Defect Detection

Influence of the inlet boundary layer on non-axisymmetric endwall contouring effects in a linear compressor cascade

Effective strategy of non-axisymmetric endwall contouring in a linear compressor cascade

Contact Info

Product

Resources

About