The through-flow method still plays an important role in the design of modern aero-engine, and its accuracy depends on the loss and deviation model. The presence of tip clearance will impact the deviation distribution, while the retained lift is somewhat related to this effect. To achieve a more precise deviation model, this paper utilises the machine learning approach. The database comprises cascades with tip clearances in training, from which obtains a span-wise deviation model and executes its validation by comparing with experiment result. The database is obtained by calculating 16 different geometries of the cascades with tip clearance in different working conditions, introducing the geometrical parameters of the cascades and retained lift as feature engineering. The deviation and the retained lift follow the same trend with tip clearance size and operating conditions variation. We predict the span-wise distribution of the retained lift using the k-nearest neighbour regression, and then combine with the traditional model to get the distribution of the deviation. The results show that the coefficient of determination of the retained lift coefficient prediction in the test set reaches 81.02%, and the mean absolute error is around 1.32%. Moreover, the trend predictions of cascade deviation distribution for different tip clearance size are all in good agreement with the experimental results. The coefficient of determination of the prediction with the simulation is 75.23%, and the mean absolute error is 1.74%.

The utilization of a part-span connector (PSC) has the potential to enhance the blade frequency, but with the penalty of aerodynamic performance. In this study, we numerically investigate the aerodynamic performance of two types of bionic structure snubbers: (1) Harbor seal whisker (HSW) and (2) Atropus’s body shape (ABS). The investigation is conducted by solving the three-dimensional Reynolds-Averaged Navier-Stokes (RANS) equations and utilizing the SST turbulent model. In this study, the performance impact of classical snubbers on a cascade blade has been examined by modeling it with and without an ellipse-shaped snubber. The vortex induced by the snubber predominantly manifests on the suction side and can be categorized into three primary vortices: upper, lower, and tail. The upper and lower vortices serve as the primary contributors to loss. Compared to the conventional ellipse snubber, the ABS snubber exhibits a reduction in the total pressure loss coefficient by 0.11% and an increase in the mass flow rate by 0.41%. On the contrary, the implementation of the HSW snubber has the potential to mitigate parameter fluctuations. However, it is important to note that this comes at the cost of a 0.10% increase in the total pressure loss coefficient and a 0.20% decrease in mass flow rate. This article further examines the factors contributing to these disparities.

The selection of loss models has a significant effect on the one-dimensional mean streamline analysis for obtaining the performance of centrifugal compressors. In this study, a set of optimized loss models is proposed based on the classical loss models suggested by Aungier, Coppage, and Jansen. The proportions and variation laws of losses predicted by the three sets of models are discussed on the NASA Low-Speed-Centrifugal-Compressor (LSCC) under the mass flow of 22 kg/s to 36 kg/s. The results indicate that the weights of Skin friction loss, Diffusion loss, Disk friction loss, Clearance loss, Blade loading loss, Recirculation loss, and Vaneless diffuser loss are greater than 10%, which is dominant for performance prediction. Therefore, these losses are considered in the composition of new loss models. In addition, the multi-objective optimization method with the Genetic Algorithm (GA) is applied to the correction of loss coefficients to obtain the final optimization loss models. Compared with the experimental data, the maximum relative error of adiabatic the three classical models is 7.22%, while the maximum relative error calculated by optimized loss models is 1.22%, which is reduced by 6%. Similarly, compared with the original model, the maximum relative error of the total pressure ratio is also reduced. As a result, the present optimized models provide more reliable performance prediction in both tendency and accuracy than the classical loss models.

The influence of the hole length-to-diameter ratio on film-cooling performance is numerically investigated for a cylindrical hole and laidback fan-shaped hole with an inlet groove. Numerical analysis of film-cooling is conducted by solving three-dimensional Reynolds-averaged Navier-Stokes equations (RANS) with a Realizable k-ε turbulence model. Rectangular and triangular grooves are applied to the inlet of cylindrical and laidback fan-shaped holes. The ratio of the hole length (L) to the diameter (D), i.e., L/D, is varied between 6–12 at blowing ratios (M) of 0.5 to 1.5 for the cylindrical hole and 0.5 to 3.0 for the laidback fan-shaped hole. For cylindrical holes with an inlet groove, the film-cooling effectiveness decreases as the L/D increases, regardless of the blowing ratio. However, in the case of laidback fan-shaped holes, the cooling performance with the length-to-diameter ratios shows different tendencies for each blowing ratio. At low blowing ratios (M=1.0), relatively high effects were observed with more than 5% increases in the effectiveness at L/D=10 and 12 compared to that of L/D=6. However, the performance is maximized at L/D=8 under high-blowing-ratio conditions (M=3.0). The cooling efficiency is enhanced up to 148% for square grooves and 124% for triangle grooves compared to those of L/D=6.

Investigating the interaction between purge flow and main flow in gas turbines is crucial for optimizing thermal management, and enhancing aerodynamic efficiency. Measuring the high-speed rotating rotor poses challenges; however, employing the pre-swirl method to model rotational effect can facilitate experimental measurements. This study evaluates the validity of the pre-swirl method for modeling rotational effects. Numerical simulations are conducted under both stationary conditions, with seven swirl ratios, and rotational conditions. The investigation focuses on the underlying mechanisms of pre-swirl and rotation. Pre-swirl and rotation impart circumferential velocity to the purge flow relative to the blade, resulting in a diminishing effect on endwall cooling. On the other hand, pre-swirl reduces the adverse pressure gradient, and the rotation generates Coriolis forces acting on the passage vortex, both contribute to an increasing effect on endwall cooling. Under pre-swirl condition, the diminishing effect is dominant, while in rotational condition, neither the diminishing nor the increasing effect exhibits an overwhelmingly dominant trend.

To achieve high-performance compressor cascades at low Reynolds number (Re), it is important to organize the boundary layer transition and separation processes efficiently and reasonably. In this study, the airfoil is focused on at a 5% blade height at the root of the orthogonal blade in the downflow passage of the high-load booster stage. The bionics modeling design is carried out for the leading edge of the original blade cascade; the response characteristics of laminar transition and separation to blades with different leading edge shapes at low Reynolds numbers are studied by using large eddy simulations combined with Omega vortex identification. The findings of this study demonstrate that bionic leading edge modeling can significantly improve the aerodynamic performance of blades at low Reynolds numbers. The blades effectively suppress the formation of separation bubbles at low Reynolds numbers and weaken or even eliminate large-scale flow separation at the trailing edge. In addition, the blades can weaken the vortex intensity on the blade surface, reduce the areas of high-velocity fluctuations, and minimize aerodynamic losses caused by turbulence dissipation. These results should serve as a valuable reference for the aerodynamic design and flow control of the high-load booster stage blade at low Re.

Energy storage technology is an essential part of the efficient energy system. Compressed air energy storage (CAES) is considered to be one of the most promising large-scale physical energy storage technologies. It is favored because of its low-cost, long-life, environmentally friendly and low-carbon characteristics. The compressor is the core component of CAES, and the performance is critical to the overall system efficiency. That importance is not only reflected in the design point, but also in the continuous efficient operation under variable working conditions. The diagonal compressor is currently the focus of the developing large-scale CAES because of its stronger flow capacity compared with traditional centrifugal compressors. And the diagonal compressor has the higher single stage pressure ratio compared with axial compressors. In this paper, the full three dimensional numerical simulation technologies with synergy theory are used to compare and analyze the internal flow characteristics. The performance of the centrifugal and diagonal impellers that are optimized under the same requirements for large-scale CAES has been analyzed. The relationship between the internal flow characteristics and performance of the centrifugal and diagonal impellers with the change of mass flow rates and total inlet temperature is given qualitatively and quantitatively. Where the cosine value of the synergy angle is high, the local flow loss is large. The smaller proportion of the positive area is the pursuit of design. Through comparative analysis, it is concluded that the internal flow and performance changes of centrifugal and diagonal impellers are different. The results confirm the superiority and feasibility of the off-design performance of the diagonal compressor applied to the developing large-scale CAES.

For most aircraft engines, inflow distortion is inevitable. Inflow distortion is known to degrade the aerodynamic performance and stable operating limits of a compressor. Tandem rotor configuration is an arrangement that effectively controls the growth of the boundary layer over the suction surface of the blade. Therefore, a higher total pressure rise can be achieved through this unconventional design approach involving the splitting of the blade into forward and aft sections. It is expected that the effect of inlet flow distortion would be more severe for a tandem-rotor design due to the greater flow turning inherent in such designs. However, this aspect needs to be thoroughly examined. The present study discusses the effect of circumferential distortion on the tandem-rotor at different rotational speeds. Full-annulus RANS simulations using ANSYS CFX are used in the present study. The performance of the rotor at a particular flow coefficient and different rotational speeds is compared. The total pressure and efficiency are observed to drop at lower mass flow rates under the influence of circumferential distortion. The loss region in each blade passage is mainly associated with the blade wake, tip leakage vortex, secondary flow, and boundary layer. However, their contribution varies from passage to passage, particularly in the distorted sector. At the lower span, the wake width is found to be higher than that at a higher span. Due to the redistribution of the mass flow, the circumferential extent reduces at a higher span. In the undistorted sector, the strength of the tip leakage vortex is significantly higher at the design rotational speed than at lower speeds. The distortion near the tip region promotes an early vortex breakdown even at the design operating condition. This adversely affects the total pressure, efficiency, and stall margin. Under clean flow conditions, this phenomenon is only observed near the stall point. At the design operating condition, the breakdown of the forward rotor tip leakage vortex is detected in four blade passages. The axial velocity deficit and adverse pressure gradient play a significant role in the behaviour of tip leakage vortex at lower rotational speeds in the distorted sector. A twin vortex breakdown is also observed at lower speeds.

Multi-fidelity simulations incorporate computational fluid dynamics (CFD) models into a thermodynamic model, enabling the simulation of the overall performance of an entire gas turbine with high-fidelity components. Traditional iterative coupled methods rely on characteristic maps, while fully coupled methods directly incorporate high-fidelity simulations. However, fully coupled methods face challenges in simulating rotating components, including weak convergence and complex implementation. To address these challenges, a fully coupled method with logarithmic transformations was developed to directly integrate high-fidelity CFD models of multiple rotating components. The developed fully coupled method was then applied to evaluate the overall performance of a KJ66 micro gas turbine across various off-design simulations. The developed fully coupled method was also compared with the traditional iterative coupled method. Furthermore, experimental data from ground tests were conducted to verify its effectiveness. The convergence history indicated that the proposed fully coupled method exhibited stable convergence, even under far-off-design simulations. The experimental verification demonstrated that the multi-fidelity simulation with the fully coupled method achieved high accuracy in off-design conditions. Further analysis revealed inherent differences in the coupling methods of CFD models between the developed fully coupled and traditional iterative coupled methods. These inherent differences provide valuable insights for reducing errors between the component-level model and CFD models in different coupling methods. The developed fully coupled method, introducing logarithmic transformations, offers more realistic support for the detailed and optimal design of high-fidelity rotating components within the overall performance platform of gas turbines.

Stall in compressors can cause performance degradation and even lead to disasters. These unacceptable consequences can be avoided by timely monitoring stall inception and taking effective measures. This paper focused on the rotating stall warning in a low-speed axial contra-rotating compressor. Firstly, the stall disturbance characteristics under different speed configurations were analyzed. The results showed that as the speed ratio (RR) increased, the stall disturbance propagation speed based on the rear rotor speed gradually decreased. Subsequently, the standard deviation (SD) method, the cross-correlation (CC) method, and the discrete wavelet transform (DWT) method were employed to obtain the stall initiation moments of three different speed configurations. It was found that the SD and CC methods did not achieve significant stall warning results in all three speed configurations. Besides, the stall initiation moment obtained by the DWT method at RR=1.125 was one period after the stall had fully developed, which was unacceptable. Therefore, a stall warning method was developed in the present work based on the long short-term memory (LSTM) regression model. By applying the LSTM model, the predicted stall initiation moments of three speed configurations were at the 557th, 518th, and 333rd revolution, which were 44, 2, and 74 revolutions ahead of stall onset moments, respectively. Furthermore, in scenarios where a minor disturbance preceded the stall, the stall warning effect of the LSTM was greatly improved in comparison with the aforementioned three methods. In contrast, when the pressure fluctuation before the stall was relatively small, the differences between the stall initiation moments predicted by these four methods were not significant.

The flow field at the inlet of compressors is generally encountered combined total pressure and swirl distortion for either aircraft engine with S-duct or gas turbine with lateral air intake. This inevitably deteriorates compressor aerodynamic performance, including not only the efficiency or pressure ratio but also the operation stability. In order to conquer this issue, appropriate measures such as integrating flow control techniques and modifying inlet or compressor design are of benefits. Due to this motivation, this article develops a full-annular two-dimensional (2D) and a partial-annular three-dimension (3D) optimization strategy for non-axisymmetric vane design. Firstly, two numerical simulation methods for evaluating performance of full-annular 2D vane and compressor with partial-annular 3D vane are developed. The swirl patterns at the inlet of a 1.5-stage axial compressor are analyzed and parametrized, and the parameterization is transferred to characterize the circumferential distribution of geometrical parameters of the vane profile. These approaches dramatically reduce computational simulation costs without violating the non-axisymmetric flow distortion patterns. Then various full-annular 2D sections at different radial locations are constructed as design space. The designed vane is reconstructed and 3D numerical simulations are performed to examine performance of the non-axisymmetric vane and the compressor with it. Also, partial annular 3D optimization is conducted for balancing compressor efficiency and stall margin. Results indicate that the designed non-axisymmetric vane based on full-annular optimization approach can decrease the vane total pressure loss under the considered inlet flow distortion, while those using partial-annular optimization achieve positive effects on compressor stall margin.

The influences of leading-edge tubercle amplitude on airfoil flow field have been analyzed at high angle of attack. The accuracy of a large eddy simulation (LES) research is validated through quantitative comparisons with corresponding experimental results. Then, a proper orthogonal decomposition (POD) analysis has been carried out based on the unsteady flow field and the fluid mechanisms of corresponding POD modes have been identified. Consequently, the influences of leading-edge tubercle amplitude have been uncovered. Since the streamwise vorticity is larger than that of small amplitude cases, the momentum transfer process at peaks is more obvious for large amplitude, leading to delayed flow separation. Both amplitude and wavelength play important roles in the generation of laminar separation bubble (LSB) at troughs. Moreover, the Karman vortex shedding process takes place at specific trough sections as pairs of periodic spatial structures exist in the dominant POD modes. The destruction of Karman vortex shedding process is strengthened along with the increase of amplitude.

The full annulus numerical research was performed on a low-speed compressor rotor to investigate the rotating instability in the tip region. The frequency spectra show the existence of rotating instability at narrow stable operating range. With the decrease of flow rate, 31 cells of flow disturbance can be found in the instantaneous flow field. The distribution of vortex suggests that the circumferential propagation of the interaction between tip leakage vortex and adjacent blade brings about these cells. The dynamic mode decomposition (DMD) method and spatial discrete Fourier transform (SDFT) were applied to obtain the circumferential mode features, and the results indicate that the rotating instability is associated with the 31 cells of flow disturbance. Then the DMD method was further applied on the pressure data from a circle and an annulus domain, so as to extract different mode components with the corresponding spatial structures, frequencies and amplitudes. The results suggest that DMD modes can display the flow feature and explore the evolution of each instability source in the tip flow field.

In order to investigate the aerodynamics of a high speed low pressure turbine works in high Mach number and low Reynold number environment, the effect of freestream turbulence (FST) on the boundary layer development on the high speed low pressure turbine under different Reynolds numbers (Re) is numerically investigated. Large eddy simulation is adopted here with a subgrid scale model of Wall Adapting Local Eddy viscosity (WALE). Cases with Re ranging from 100 000 to 400 000 under an exit Mach number (Ma) of 0.87 have been considered at low and high FST levels. A low Ma case (0.17) under very low Re has also been studied under both low and high FST. It is found that higher Re or FST level leads to earlier transition. Re has a greater effect than FST on the development of boundary layer. The effect of FST on the boundary layer depends on the Re. The boundary layer development shows totally different behaviors under different Ma. A separation bubble could be formed under low Ma while no attachment could be detected under high Ma. The FST has a stronger effect on the separated boundary layer under low Ma, which could eliminate the separation in the present study. For all the cases under low FST, the Kelvin-Helmholtz instability is the dominate mechanism in the transition process. For the low Ma case with high FST, the streamwise streaks play a dominant role in the transition process. For the high Ma cases with high FST, both the streamwise streaks and Kelvin-Helmholtz instability work in the transition process. The streamwise streaks play a more important role when the Re increased.

The effects of root fillet on the flow behavior of high loading compressor rotor tends to be much more crucial in practice, and it’s necessary to explore the internal relations between the geometric effects of root fillet and the flow behaviors of rotor blade. Therefore, eight types of root fillet with different radius were designed and installed around the blade root of NASA Rotor67. With the aids of fillet, the corner separation near suction side of blade root has been suppressed significantly in that the root fillet reconstructs the circumferential bending distributon of the suction-side curve from leading edge to trailing edge, and reduces the genmetric turning angle in the latter part of root section near trailing edge. However, apart from the improvement of corner flow characteristic caused by root fillet, both the tip flow deterioration and the decrease of stall margin occur in the new rotors, which indicates an indirect correlation between tip flow characteristic and root fillet exists indeed in the three-dimensional flowfields of transonic rotor. Actually, by means of the new radial pressure equilibrium affected by root fillet, a larger radius of root fillet contributes to much larger blade loading and stronger leakage flow in tip region of compressor rotor. As a result, a monotonic decrease of stall margin was present in the transonic rotor with increase of the root fillet radius. Subsequently, the positive bending of blade tip was introduced to deal with the negative effect caused by the root fillet indirectly. Combined with the effects of root fillet and positive tip-bending on the radial pressure equilibrium existing in channels, both the radial and streamwise loading distributions tend to be much more reasonable in new rotors, and the static pressure difference in former 1/3 chord of blade tip has decreased clearly which benefits to reduce the strength of leakage flow in tip region. Therefore, the flow deterioration in tip region of transonic rotor induced by root fillet has been well suppressed, with an obvious improvement of overall performance occurring in new rotors.

Hyperloop has become one of the key reserve technologies for future high-speed rail transit. The gas in the tube is compressed and rubbed, leading to a strong aerodynamic heating effect. The research on the flow field characteristics and aerodynamic heating effect of hyperloop is in its infancy, and that on the flow field structure is lacking. In this study, the nozzle theory was used to make a preliminary judgment on the choked flow phenomenon in the hyperloop. Based on the flow results obtained under different working conditions, the identification basis of the choked flow phenomenon in the hyperloop was obtained. Furthermore, the effect of the choked/unchoked flow on the flow structure, temperature, and pressure distribution of the annular space in the tube was analyzed. Based on traditional high-speed railway aerodynamics, according to relevant theories and calculation in aerospace field, and combined with the model test data, the reliability verification analysis on the characteristics of the flow field are carried out. The structure of the flow filed in the tube can be divided into choked and unchoked. The judgment is dependent on whether the throat reaches the speed of sound. Under the choked flow, a normal shock wave is formed in front of the tube train. The temperature rise of the local flow field exceeds 50 K; the temperature rise of the stagnation region exceeds 88 K, and the pressure is approximately 1.7 times that of the initial pressure in the tube. When the flow is unchoked, differences arise in the distribution of the flow field corresponding to different incoming Mach numbers. When the incoming flow is supersonic, the flow field maintains a supersonic speed, and a bow-shaped shock wave is formed at the front of the tube train. Owing to the shock wave or expansion wave, the local flow field exhibits significant fluctuations in temperature and pressure. Conversely, when the incoming flow is subsonic, the flow field in the tube maintains a subsonic speed, and no shock wave structure is observed.

To overcome the huge drag on an airfoil in transonic flow, a hybrid flow control method using suction and loaded leading edge (SLLE) is proposed and its active feedback control effect is studied in the different operation conditions. The loaded leading edge structure can redistribute the pressure as a passive flow control technique, while the suction slot is used to control shock wave’s position and flow separation, which can be conducted actively and automatically using feedback control system. Firstly, the investigation is conducted in steady flow, and a significant drag reduction performance is obtained. The highest drag reduction rate of 22.5% can be got when attack angle is 5°, and the increasing of lift-drag ratio can be obtained in each attack angle case. Secondly, a heuristic approach to feedback flow control is conducted in off-design inflow conditions, where a feedback-based SLLE control method is introduced. The results show the SLLE control can achieve a fair drag reduction performance which is over 10%, which indicates to a flow control method with good applicability in changing flow conditions.

Cavity-based flameholder is expected to be applied for ramjets or afterburners, which could work efficiently in the high-altitude space with low pressure. The detailed fluid structure helps to understand the flame stability principle of the flameholder. The fluid structure in the center section and side section of the cavity-based flameholder is experimentally measured at the inlet pressure of 0.04–0.10 MPa, Mach number of 0.1, and temperature of 300 K. Results indicate that the inlet pressure has a significant effect on the fluid-structure in the cavity. The bluff body affects the generation of the vortex in the cavity. As the inlet pressure decreases from 0.10 MPa to 0.04 MPa, the classical dual-vortex maintains excellent stability in the side section of the cavity. Whereas the single-vortex in the center section gradually becomes incomplete with the inlet pressure varying from 0.10 MPa to 0.06 MPa, and it disappears at 0.04 MPa. The reason is that with the reduction of inlet pressure, the density decreases as well, and the proportion of the mass flow rate attracted to the low-pressure area downstream of the bluff body increases, which leads to the vortex being gradually pulled and destroyed.

The inlet swirl distortion and non-uniform tip clearance have great effects on aero-engine performance and stall margin. In this paper, the effects of paired swirl distortion on the aerodynamic stability and stall inception of a single stage axial compressor with non-uniform tip clearance are quantitatively analyzed by using the swirl distortion descriptors. The experimental results show that the paired swirl distortion dominated by co-rotating swirl improves the stability of the axial compressor. For a single-stage axial compressor with eccentricity of 100%, the stall inception starts at the maximum tip clearance with clean inlet. The initial position of the stall inception is determined by the maximum tip clearance when the small intensity paired swirl distortion exists at the compressor inlet. As the swirl intensity increases, it shifts towards the position of the counter rotating swirl vortex core. The inlet swirl will not change the type of stall inception.

Transonic tandem cascades can effectively increase the working load, and this feature conforms with the requirement of the large loads and pressure ratios of modern axial compressors. This paper presents an optimization strategy for a German Aerospace Center (DLR) transonic tandem cascade, with one front blade and two rear blades, at the inlet Mach number of 1.051. The tandem cascade profile was parameterized using 19 control parameters. Non-dominated sorting Genetic algorithm (NSGA-II) was used to drive the optimization evolution, with the computational fluid dynamics (CFD)-based cascade performances correction added for each generation. Inside the automatic optimization system, a pressure boundary condition iterative algorithm was developed for simulating the cascade performance with a constant supersonic inlet Mach number. The optimization results of the cascade showed that the deflection of the subsonic blade changed evidently. The shock wave intensity of the first blade row was weakened because of the reduced curvatures of the optimized pressure and suction sides of the front blade part and the downstream moved maximum thickness position. The total pressure losses decreased by 15.6%, 20.9% and 19.9% with a corresponding increase in cascade static pressure ratio by 1.3%, 1.8% and 1.7%, for the three cascade shapes in the Pareto solution sets under the near choke, the design and near stall conditions, respectively.

Flow instability in the centrifugal compressor should be detected and avoided for stable and safe operation. Due to the popularity of electric centrifugal compressors, instability detection could be achieved by measuring motor signals instead of traditional aerodynamic signals. In this paper, the feasibility of instability detection by motor signals (i.e. rotating speed and phase current) was studied experimentally. The physical structure and control method of the electric centrifugal compressor were discussed to reveal the potential of instability detection by motor signals. Dynamic pressure signals and motor signals measured during unsteady experiments were analyzed in the time domain and frequency domain. Characteristics of these signals were then compared under different operating conditions to indicate the feasibility of instability detection by motor signals. Finally, the ability of Short-Time Fourier Transform (STFT) of rotating speed signals in real-time instability detection was discussed. Results showed that the rotating speed signal is a good alternate for instability detection in spite of signal distortion, while the phase current signal can only detect surge due to the low resolution of the controller. Based on the variations of the amplitude and frequency of rotating speed signals, the real-time instability can be captured accurately by STFT with a window size of 0.5 s. Besides, the interference caused by the controller can be removed by STFT.

This paper proposes a new-designed rim seal configuration with sealing holes based on the conventional radial rim seal, and presents a numerical comparison of the sealing performance between the conventional sealing flow supply configuration and the new sealing flow supply configuration with holes at different sealing flow rates. The sealing effectiveness and unsteady flow yields at the rim seal are numerically simulated by using the URANS method and SST turbulent model from ANSYS CFX. The influence of the new sealing flow supply configuration on the sealing effectiveness at different sealing flow rates is determined. The effectiveness of different sealing flow rates in the conventional rim seal is also studied. As to the conventional rim seal, the increase in the sealing flow rate reduces the degree of gas ingestion induced by the effect of mainstream ingress at the rim clearance, while the unsteady flow characteristics are enhanced, and the number and amplitude of the low-frequency signals increase. The position of the Kelvin-Helmholtz instabilities vortex structures is left by the increased sealing flow rate, and its strength is suppressed. Compared with the conventional rim seal configuration, the new sealing flow supply configuration with holes could reduce the sealing efficiency by 5.06% at most at sealing flow distribution m₁:m₂=3:1 when C_w=2000, and improve the sealing efficiency by 11.71% at most at sealing flow distribution m₁:m₂=1:1 when C_w=7500. It shows that the lateral jet from the holes induces a larger-scale Kelvin-Helmholtz vortex structure at C_w=2000, thus the sealing efficiency in the wheel space is also reduced. However, the size of the Kelvin-Helmholtz vortex structures is significantly suppressed by the new sealing flow supply configuration at C_w=7500, which is beneficial to improving the sealing effectiveness of the conventional rim seal.

A Sequential Approximate Optimization framework (SAO) for the multi-objective optimization of lobed mixer is established by using the BP neural network and Genetic Algorithm: the ratio of lobe wavelength to height (η) and the rise angle (α) are selected as the design parameters, and the mixing efficiency, thrust and total pressure loss are the optimization objectives. The CFX commercial solver coupled with the SST turbulence model is employed to simulate the flow field of lobed mixer. A tetrahedral unstructured grid with 5.6 million cells can achieve the similar global results. Based on the response surface approximation model of the lobed mixer, it is necessary to avoid increasing or decreasing α and η at the same time. Instead, the α should be reduced while the η is appropriately increased, which is conducive to achieving the goal of increasing thrust and reducing losses at the expense of a small decrease in the mixing efficiency. Compared with the normalized method, the non-normalized method with better global optimization accuracy is more suitable for solving the multi-objective optimization problem of the lobed mixer, and its optimal solution (α=8.54°, η=1.165) is the optimal solution of the lobed mixer optimization problem studied in this paper. Compared with the reference lobed mixer, the α, β (the fall angle) and H (lobe height) of the optimal solution are reduced by 0.14°, 1.34° and 3.97 mm, respectively, and the η is increased by 0.074; its mixing efficiency is decreased by 4.46%, but the thrust is increased by 2.29% and the total pressure loss is decreased by 0.64%. Downstream of the optimized lobed mixer, the radial scale and peak vorticity of the streamwise voritices decrease with the decreasing lobe height, thereby reducing the mixing efficiency. For the optimized lobed mixer, its low mixing efficiency is the main factor for the decrease of the total pressure loss, but the improvement of the geometric curvature is also conducive to reducing its profile loss. Within the scope of this study, the lobed mixer has an optimal mixing efficiency (ε=74.14%) that maximizes its thrust without excessively increasing the mixing loss.

To optimize the aerodynamic performance of the automobile cooling fan (ACF), the internal flow field of the original fan was numerically simulated. According to the theory of boundary vorticity dynamics (BVD), the distribution laws of the boundary vorticity flux (BVF) on the blade surface and the circumferential vorticity (CV) at the wake plane of the fan were analyzed, and the underlying various negative factors, such as vortex shedding, separated flow and complicated secondary flow, on the fan blade surface and its dynamic source were diagnosed. Combined with the velocity triangle theory, the mathematical relationship between the BVF diagnosis and the geometrical characteristics of the blade profile (hereinafter referred to as profile) is used to guide the design improvement of the blade. The analysis found that at the same speed, the extension and rotation of the profile could match a smaller input torque at the same flow rate and pressure rise, thereby improving the efficiency of the fan. The test results confirmed the above conclusion. The peak efficiency of the improved fan has been increased by 2.3%, and the aerodynamic performance in the low-flow-rate has been improved. The conclusion of the study shows the applicability of the BVD theory in the diagnosis and design improvement of ACF internal flow.

Film cooling effectiveness superposition of double-row injection holes on the turbine vane was studied by infrared temperature measurement experiment. The Sellers superposition method and a modified Sellers method were adopted for dustpan-shaped hole and cylindrical hole. Numerical simulations were implemented to analyze the film superposition mechanism. It is found that the Sellers method is more accurate on the suction side than the pressure side. Injection film of the two types of holes exhibits different superposition modes. Cylindrical hole are “blocky-like” superposition. Dustpan-shaped hole are “sheet-like” superposition. The counter-rotating vortex pairs and separation of the film are the main factors affecting the accuracy of Sellers film superposition method. The modified method can significantly improve the superposition prediction accuracy for almost all situations. The modified method reduces superposition errors from 28% to 3% for the cylindrical hole, and from 42% to 13% for the dustpan-shaped hole on the suction side. It reduces superposition errors from 30% to 8% for the cylindrical hole, and from 23% to 15% for the dustpan-shaped hole on the pressure side.

In the traditional design of the centrifugal compressor, the splitter blade and the main blade always keep the same shape. However, to enable high efficiency of the high-loading centrifugal compressor, the matching of design parameters of the splitter blade and the main blade needs to be optimized. In this paper, the influence of the load distribution between the main blade and the splitter blade on the aerodynamic performance, the flow field, and the internal vortices of a high-loading centrifugal compressor were studied by means of CFD prediction. Four cases with different values of the variable CR which is defined as the load-ratio of splitter blade to main blade were set up. In each case, the splitter blade and the main blade were shaped according to different laws of circulation distribution (rV_u) while the average circulation of the splitter blade and the main blade at any meridional position were consistent with that of the prototype. The results showed that a proper reduction of the load-ratio of splitter blade to main blade is beneficial to suppress the leakage vortex of the splitter blade and reduce the scale of the wake in the channel near the suction-side of the splitter blade, which consequently improves the flow uniformity at the impeller outlet and enhances the aerodynamic performance of both the stage and the component. The stage isentropic efficiency of the optimal case was found to be 0.7% higher than that of the prototype and the stage total pressure ratio was also improved. The optimal value of CR, which in this investigation is 94%, is supposed to be the result of the trade-off between the development of the wake and the leakage vortices in adjacent two channels. The optimization of the load distribution between the main blade and the splitter blade provides an opportunity to further improve the high-loading centrifugal compressor performance.

Interaction between the coolant and the secondary flow plays an important role in endwall cooling performance. For the leading-edge region, oncoming main flow inside the boundary layer impinges onto the vane leading edge and turns into the horseshoe vortex. Horseshoe vortex entrains coolant off the surface, thus posing severe challenges to the cooling design there. Based on analyses on the leading-edge vortex formation mechanism, a new kind of endwall film cooling design, vertical hole upstream of the saddle point, is proposed to obtain more uniform film coverage over the vane/endwall junction region. Coolant injected from the vertical hole can pass over the horseshoe vortex and impinge around the stagnation line on the vane leading edge. Uniform film coverage can be obtained around the vane leading edge where coolant clings to the endwall surface due to the span-wise pressure gradient of the stagnation region. Numerical simulations are conducted about the cooling performance of two main kinds of both isotropic and anisotropic hole geometries for the endwall and vane surface. Results come that the anisotropic hole shows significant advantages over the isotropic one because it suppresses the symmetrical kidney vortices thus weakening the mixture with high-temperature gas. Blowing ratio (M) effect is analyzed and conclusions are drawn that the cooling performance of the endwall around the leading edge is sensitive to M and adiabatic film cooling effectiveness peaks at about M = 2.0. Better cooling performance over the vane corner region can be obtained when M gets even higher while the effective film coverage area shrinks. Apart from that, the phenomenon of phantom cooling on the upper triangular region of the suction surface can be observed when coolant on the endwall is entrained by the vortex formed at the corner of the leading edge.

Theoretical and numerical study was carried out based on a linear turbine cascade (the Basic cascade) to compare the influences of the increased cascade pitch and turning angle in this paper. On one hand, the two highly-loaded designs both reduced the stability of flow field through enhancing adverse pressure gradient and span-wise pressure gradient of the fluid near suction surface. Therefore, the two highly-loaded designs would both result in thicker boundary layer and stronger secondary flow, so the secondary loss would be increased and more difficult to suppress in the highly-loaded cascades. On the other hand, the two highly-loaded designs showed different influences on the pitch-wise migration of the fluid near the endwall (cross flow) because of the different load enhancing mechanisms. In other words, the increased cascade pitch (TCx highly-loaded design) would delay the pitch-wise migration of the horseshoe vortex because of the increased channel width, while the increased turning angle (Turn highly-loaded design) would do the opposite because of the increased pitch-wise pressure gradient. As a result, the enhancement of the interaction between the fluid near the suction surface and the cross flow would be much stronger in the Turn highly-loaded design than the TCx highly-loaded design, and the span-wise developing tendencies of vortexes and fluid near the suction surface would show much stronger enhancing tendency in the former than the latter.

To achieve efficient control of supersonic compressor cascade flow, a type of spanwise distributed pulsed arc discharge plasma actuation (PADPA) was designed. To simulate the influences of PADPA on the flow field, a phenomenological model was established. Then, the flow control effects of PADPA on supersonic compressor cascade flow were researched numerically. The results show that under low static pressure ratio condition, the compressive wave induced by PADPA reduced the intensity of the passage shock wave, which eventually reduced shock wave loss. It was also found that PADPA produced an adverse pressure gradient (pre-compression effect) around the actuation location, which reduced the strength of the high adverse pressure gradient induced by the passage shock wave. The airflow on both sides of the actuation location was accelerated by PADPA owing to the spanwise distributed layout. Thus, it improved the ability of the boundary layer to resist the effect of the adverse pressure gradient and reduced the separation zone. Consequently, the total pressure loss was reduced by 6.8%. Under high pressure ratio condition, the effect of PADPA on the suction side controlling the large separation of the boundary layer was insignificant. The total pressure loss also increased slightly.

The present study focuses on the influence of the swirling flows on flow behaviors and performance of a radial-flow turbocharger turbine under pulsating inflow condition. To characterize the effects of swirling flow, three sets of simulations of the turbine were carried out, which are an unsteady simulation under pulsating swirling inflow, an unsteady simulation under equivalent pulsating uniform inflow, and quasi-steady simulations under uniform inflow. Results proved that swirling flow has a considerable negative influence on turbine instantaneous performance and lead to 2.5% cycle-averaged efficiency reduction under pulsating flow condition. Swirling inflow would lead to significant losses in both the volute and the rotor, while the pulsating inflow leads to higher losses in the rotor and shows little influence on the losses in the volute. The instantaneous efficiency reduction of the turbine could be correlated with the time-varying inlet swirl strength. Under the influence of unsteady inlet swirls, the volute flow field is highly distorted and the free vortex relation is no longer valid. The swirling flow has strong interactions with the wake flow of the volute tongue, leading to additional losses. Relative flow angle at rotor inlet is remarkably reduced and its distribution is significantly distorted. Strong separation flows and passage vortices would appear in the rotor because of the swirling inflow, leading to inferior rotor performance.

In a variable nozzle turbine (VNT), the nozzle vane and rotor clearance can generate complicated tip leakage flow, which produces a large flow loss and results in a noticeable reduction in the VNT performance. Therefore, it is necessary to study the influence of the nozzle vane and rotor clearance on the VNT performance to reveal the underlying mechanisms. In this study, the variation of VNT performance resulting from the nozzle vane and rotor clearance was studied numerically using the commercial CFD code. The numerical results were in good agreement with the experimental data. VNT performance under three varied clearance conditions under different clearance sizes, rotating speeds, and expansion ratios was analyzed. Clearance conditions included the individual effect of nozzle vane clearance, the individual effect of rotor clearance, and the combined effect of nozzle vane and rotor clearance. The results showed that the rotor clearance (clearance size equal to 0.4 mm) resulted in a certain reduction in thermal efficiency (–3.98%) and torque (–4.48%). Meanwhile, the nozzle vane hub and shroud clearance also affected the thermal efficiency (–3.66% and –13.37%, respectively) and torque (12.54% and –0.05%, respectively). Compared with the rotor clearance, the nozzle vane clearance dominated the variation in the VNT performance, and the values of the mass flow ratio, thermal efficiency, and torque increased by 12.15%, 5.43%, and 8.01%, respectively (The clearance on both sides of the vane was equal to 0.2 mm.). Under the combined effect of the nozzle vane and rotor clearance, the deviation of the thermal efficiency and mass flow ratio was smaller than the sum of the values of their individual effects; when the clearance on both sides of the vane was equal to 0.2 mm and the rotor clearance was equal to 0.6 mm, the value was reduced by 2.77% and 1.71%, respectively. This study also explains the influence mechanisms of clearance at both ends of the nozzle vane and the vane/rotor clearance interaction on the VNT performance in detail.

This paper presents a novel approach of modeling the air-cooled turbine with CFD-based throughflow analysis. Starting from the basic equations of motion, governing equations and source terms for mass, momentum and energy are formulated in an analytical manner. These source terms are to mimic the authentic injection-mainstream interactions with easy implementation. The source terms in the aero-cooling scenario are related to corresponding sources in the aerodynamic-only analysis. Based on such formulations, a novel strategy is developed to estimate aerodynamic characteristics of a blade row under film cooling with known characteristics under no cooling. The model and the strategy are validated in the classic NASA E3 turbine guide vane under various operating conditions. Sensitivity studies of input parameters are conducted to evaluate the applicability of the proposed model. Specifically, the flow rate distributions of cooling flow at different cooling holes are crucial for accurate predictions.

In this paper, the aero-thermal performance of squealer tips with deep-scale depth is numerically investigated in an axial flow turbine, which is compared with the squealer tip with traditional cavity depth. Numerical methods were validated with experimental data. The effect of cavity depth and tip clearance was considered. The numerical results show that for the squealer tip with conventional cavity depth, the size of the reflux vortex enlarges as the cavity depth increases. The velocity and uniformity of high entropy production rate (EPR) inside the cavity reduce obviously with the cavity developing into deep-scale. However, the increase of depth 10% of the blade span (H) leads to enlargement of cavity volume, which increases the total entropy production rate. And the overall dimensionless entropy production rate (DEPR) of gap and cavity obtains a maximum increase of 43.54% in contrast to the case with 1%H depth cavity. As a result, the relative leakage mass flow rate reduces by 20.6% as the cavity depth increases from 1% to 10%. Given the heat transfer, as the cavity significantly increases to 10%H, the enhanced cavity volume results in a more enormous cavity vortex with low velocity covering the floor, which weakens the convective heat transfer intensity and reduces the area of high heat transfer. The normalized average heat transfer coefficient at the cavity bottom reduces by 40.26% compared to the cavity depth of 1%H. In addition, the deep-scale cavity is more effective in inhibiting leakage flow at smaller tip clearance. The reduction amplitude of normalized average heat transfer coefficient at the squealer floor decreases as tip clearance increases, which reduces at most by about 72.6% for the tip clearance of 1%H.

Nonaxisymmetric endwall is an effective method to reduce secondary loss and improve aerodynamic performance. In this paper, a nonaxisymmetric endwall automated optimization process based on the nonuniform rational B-spline surface (NURBS) technique was proposed. This technique was applied for the aerodynamic optimization of the turbine stator shroud endwall to reduce total pressure loss and secondary kinetic energy. The flow fields of the datum endwall design (Datum) and optimization endwall design (Opt) were investigated and compared. Quantitative loss analysis was performed with a loss breakdown method. The entropy generation was classified as profile loss, secondary loss and trailing edge loss, all of which were reduced. The secondary loss was much smaller than the profile loss. In general, the blade row total entropy loss decreased by 11.7%. The results showed that the Opt design reduced total pressure loss and coefficient of secondary kinetic energy by 11.1% and 11.0%, respectively. The decrease in secondary kinetic energy could be attributed to the reduction in the horseshoe vortex and the reduced transverse pressure gradient. When the outlet Mach numbers and inlet incidence angles vary, the performance of the profiled endwall design was always better than the datum design. In the turbine stage simulation, the efficiency was increased by 0.28% with nonaxisymmetric endwall.

In order to reveal the different effect mechanisms of blade sweep on the aerodynamic performance when a transonic rotor operates alone or in fan stage environment, two series of forward and backward swept rotors were designed and utilized in the first stage of the dual-stage NASA CR-120859 fan. Results show that, the influence of sweep on the single rotor and the whole stage is different, indicating swept designs for rotor alone may not be suitable for the stage operations. The distinct effect of sweep is account for the difference of the flow field characteristic and stall mechanism of the single rotor and the rotor in the stage environment. The single rotor is tip limited and its stall mechanism is shock/tip leakage vortex (TLV) interaction, whereas the fan stage is hub limited and its stall mechanism is the severe corner separation at stage hub region. For the single rotor, forward sweep increases the stall margin (SM) for all sweep schemes, while backward sweep reduces it in general. For the fan stage, however, backward swept rotor significantly increases stall margin and the stall mechanism is changed to shock/TLV interaction. On the contrary, forward sweep reduces stall margin in general. The flow mechanism is that forward sweep reduces blade loading at tip region near leading edge (LE) and causes the shock to move downstream. Both the variations improve flow field at tip region, while backward sweep exerts an opposite effect. At hub region, backward sweep reduces radial flow tendency by varying radial pressure gradient, causing reduction of corner separation at rotor hub, while forward sweep enhances corner separation. Moreover, with increasing of swept height and swept angle, the chock mass flow, peak efficiency and total pressure ratio of forward sweep are reduced in general, while an opposite effect can be found for backward sweep.

As continuous of the previous sand-dune-inspired design, the Barchan-Dune-Shaped Injection Compound (BDSIC)’s film cooling performance at the endwall region was further investigated both experimentally and numerically. While the public 777-shaped hole was served as a baseline, the BDSIC’s endwall effectiveness was assessed at various blowing ratios. Experiments were performed in a single-passage transonic wind tunnel using pressure-sensitive paint (PSP) technique. Carbon dioxide was used as coolant with density ratio of DR=1.53. The purge slot’s blowing ratio was fixed at M=0.3, but the coolant holes were adjusted within M= 0.5–2.0. The measured experimental results indicate that the film distribution at the endwall is strongly affected by the passage flow structures. The BDSIC holes demonstrate much higher film effectiveness than the 777-shaped holes for all blowing ratios, ~30% enhancement for regionally averaged effectiveness at M=1.0 and ~26% at M=2.0. As shown by the numerical results, the existence of BDSIC reduced the coolant penetration effect at a higher blowing ratio. Coolant was deflected and its momentum increased in the streamwise direction, therefore providing more robust film coverage over the endwall region. The anti-counter-rotating vortex pair induced by the BDSIC further stabilized the coolant film and increased the coolant spreading downstream.

A compressor test platform was designed in the purpose of easy assembly and cost-saving tests. New design concepts were firstly used on the test platform and iterative revisions were conducted to verify the effects, which decrease the risk of applying to the full annulus directly. The platform was a sector cascade that can be inserted into an otherwise full stator annulus, with a set of exchangeable endwall casing and blades manufactured by 3-D printing. The platform can create an operating condition which is closer to engine-realistic flow conditions than traditional cascade tests. The flow field of the prototype stator was tested in detail at the operating point and showed some flow defects in the tip region, and then three design plans were conducted to experimentally investigate which design concept could improve the blade tip flow. The concept of casing endwall profiling was the most effective one which markedly depressed the separation of the tip region and decreased the total pressure loss coefficient. The test results show that the design of the test platform was successful and promising, which could be used to conduct more researches on the flow mechanism of the middle stage.

The gas turbine is the main power equipment for naval ship and special civil ship, while the compressor is one of the core structures of the gas turbine. The existing tip clearance could prevent the compressor blade and casing collision. Therefore, the flow loss in the tip region caused by the tip clearance will degrade the performance of the compressor. To improve the variable clearance characteristics of the high subsonic compressor cascades, the cascades with tip clearances of 1%, 2% and 3% chord length are studied through experimental measurements and numerical calculations. The research results prove that the pressure surface tip winglet can cause a significant improvement effect under most working conditions. If the blade tip clearance size is gradually increasing within a reasonable range, the improvement effect becomes more remarkable, and the optimal tip winglet case changes. When tip clearance is 1% chord length, the PTW1.0 case (the width of the pressure surface tip winglet is 1.0 time of the original tip) reduces the flow loss by 3.09% compared with the NTW case (No Tip Winglet). When tip clearance is 2% chord length, the flow loss of PTW1.5 case (the width of the pressure surface tip winglet is 1.5 times of the original tip) is reduced by 3.46%. When tip clearance is 3% chord length, all alternative tip winglets reduce the total pressure loss, and PTW2.0 case (the width of the pressure surface tip winglet is 2.0 times of the original tip) is the best choice, which has a 6.53% degree of improvement.

Flow control methodologies have been widely used to eliminating flow separation and increasing the blade load in axial compressor. Aiming at revealing the flow mechanism of coupled bowed blading and boundary layer suction in a supersonic compressor cascade, a cascade with a diffusion coefficient of 0.62 is numerically presented. First of all, according to the available experimental data, the numerical method was validated; then, different bowed blading effects on flow field in detail were investigated; at last, based on the flow physics of purely bowed blading, the positively bowed blade was coupled with boundary layer suction on blade suction surface, whereas the negatively bowed blade was coupled with endwall suction. For coupled control method, influence mechanism on flow field, especially on the shock structure was revealed, and different aspect ratios of coupled control method were investigated as well. Results showed that the coupled positively bowed blading and suction surface suction can eliminate the flow separation effectively. Compared with that of baseline supersonic cascade, the total pressure loss coefficient of the coupled scheme was reduced by 37.4% at most. At mid-span, the shock moved downstream and the single shock was separated to a dual-shock structure since the positively bowed blading reduced the static pressure of mid-span. The coupled negatively bowed blading and endwall suction also effectively enhanced the performance of cascade by removing the corner separation, with the loss coefficient reduced by as much as 41.9%. However, the suction coefficient of optimal coupled negatively bowed blading scheme reached 10.5%, which is too high for practical use. After coupled control, the 3D shock structure became “C” shaped distribution along spanwise because of the difference in influence mechanism of negatively bowed blading on different spanwise location. Due to the opposite influence effect of positively and negatively bowed blading, the shock structure in the two different schemes of cascades were different and showed opposite variation trends as aspect ratio increased.