Abstract The aerodynamic performance of turbine components constituting the gas turbine engine is seriously required to be improved in order to reduce environmental load. The energy recovery efficiency in turbine component can be enhanced by the increase of turbine blade loading. In this study, as the first stage to investigate the aerodynamic performance of an ultra-highly loaded turbine cascade (UHLTC) with a turning angle of 160 degrees at transonic flow regime, two-dimensional steady compressible flows in UHLTC were analyzed numerically by using a commercial CFD code to focus on the profile loss. In the computations, the isentropic exit Mach number was varied in the wide range from 0.3 to 1.8 in order to examine the effects of exit Mach number on the shock wave formation and the associated profile loss generation. The computed results were examined in detail by comparing with those for a typical transonic turbine cascade. The detailed examination for the present computed results clarified the variation of shock pattern with the increase of exit Mach number and the loss “plateau” behavior in the present UHLTC.

Abstract Blade loading on a single stage high pressure centrifugal compressor is limited due to separation that might occur on the suction side of the airfoil at mass flow rates lower than design point. A novel configuration of centrifugal compressor is designed and analyzed to overcome this issue by placing multiple rotors on the same hub with a stator vane in between similar to a multi-stage axial compressor blade arrangement. By having independent rotors, blade loading can be distributed more efficiently and higher pressure rise can be achieved through this design. As the blade chord length is reduced due to splitting of single impeller blade, the effective turning angle is divided through several stages thereby lowering the adverse pressure gradient reducing the chance of separation. Stator vanes are placed in between the rotors so that the successive rotor receives the flow at desired incidence angle. The attempt here is to apply the same principle of axial compressor multi-staging on a centrifugal compressor and compare the performance with single stage using low to high fidelity analysis framework developed in-house. A low fidelity 1D analysis tool CIMdes is used for evaluating blade angles and stage degree of reaction which are exported to T-blade3, in-house parametric geometry tool, for 3D blade generation. These blades are further analyzed using 3D CFD analysis using an in-house automated multifidelity framework. Loss quantification revealed that diffuser losses are higher in singlestage and the novel design increased the backsweep angle resulting in lower diffuser losses. Splitting the single rotor facilitated the increase in backsweep angle to a larger range as compared to single rotor impeller configurations. Two configurations with different shroud height for the single stage compressors are investigated and compared with the novel compressor with respective flowpaths at 100% speedline using a multi-fidelity design analysis suite. The flow capacity is extended near the stall with a penalty in efficiency for configuration-1. Configuration-2 showed an improvement in efficiency at design mass flow rate. The preliminary analysis demonstrates the advantages of the multi-staging on the same hub and extends the design space for performance range improvement with some trade-offs.

Abstract Performance prediction and blade generation in a preliminary design stage of centrifugal compressors is critical to have a successful design. In this paper, a one-dimensional meanline design and analysis tool has been developed for single-rotor and novel multi-rotor centrifugal impellers. Loss models used for performance prediction of single stage compressors have been extended to single hub multi-rotor compressors to evaluate isentropic efficiency and pressure ratio. Stage conditions like work ratio and stator turning angle are given as input parameters and the tool computes flow properties along the meanline and also, generates velocity triangles, streamlines, smooth definition of blade angles at different spanwise sections. The tool acts a preprocessor for Tblade3 which is an in-house 3D parametric blade geometry generator to create blades. A 0D tool has been developed for multi-rotor impellers to provide an estimate of work ratio. 0D coupled with 1D tool can provide a good preliminary design point. The process of 0D to blade generation has been automated enabling it to connect with high-fidelity analysis. The motivation to create this tool is to calculate flow angles, metal angles, velocity triangles, ease of parametric modification and reduce aero-design cycle time. This tool is modular which adds the flexibility of capability extension. The code is validated with DLR centrifugal compressor experimental data. The 1D tool is also used to calculate performance and blade angles for the novel single hub multi-rotor centrifugal compressor demonstrating the versatility of the low fidelity tool. The tool suite is written in Python and is open source https://github.com/msaisiddhartha/CIMdes.

Nowadays, the flow field at the compressor is more and more complex with the increasing of the aerodynamic loading. The complex flow in the endwall regions is thus key to aerodynamic blockage, loss production, and finally its performance deterioration. The design of Blended Blade and End Wall (BBEW) contouring technology had been proved to be useful in delaying, reducing, and eliminating the corner separation in the compressor. The BBEW technology can adjust the dihedral angle between the suction and the endwall in 30% of the spanwise easily, which is different with the fillet. However, the design of the BBEW always relies on the experiences of the designers, and the effective design results cannot be the optimal result. This paper presents an optimization design method for the BBEW technology, and analyses the flow mechanism of the BBEW design. Firstly, the parameters for the BBEW design is simplified as two, one is the maximum blended width, the other is the axial position of the maximum blended width. The optimal result can be obtained through the response surface method. Secondly, based on the optimization method, this paper make an optimization BBEW design at the suction side of a NACA65 linear compressor cascade with the turning angle 42 degrees. The numerical results show that the optimal BBEW design can eliminate the boundary layer separation at the corner intersection region, and reduce the suction side separation. When the incidence angle is 0 degrees, the BBEW technology can reduce the total pressure loss coefficient by 5%, and reduce the aerodynamic blockage coefficient by 14%. The aerodynamic performance of the cascade shows a more obvious improvement with the BBEW design at a larger incidence. The total pressure loss coefficient of the cascade is reduced by 20% at 15 degrees incidence. The numerical study shows that the design with the BBEW can control the axial development of the dihedral angle between the suction side and the endwall, which can eliminate the boundary layer separation at the corner intersection region. What’s more, the BBEW technology can produce a pressure gradient at the axial position of the maximum blended width, and value of the pressure gradient in proportion to the maximum blended width. This pressure gradient enhance the kinetic energy of the low energy fluid at the endwall region, which is consist of the secondary cross flow, thus elevating the capability to withstand the adverse pressure gradient, and improve the suction side separation around the trailing edge.

High-pressure ratio is one of the important characteristics of the sustainable development of the modern aero-engine compressor components. When the fluid flows through the compressor cascade row, it will be influenced by both the streamwise pressure gradient and the transverse pressure gradient, which will cause hub-corner separation or stall. In this paper, different diffusion factors are chosen for the cascades. Each diffusion factor has different turning angles. The formation mechanism of hub-corner separation is studied under the condition of zero angle of attack. Numerical simulation is used to study the influence of pressure gradient on the flow field in the corner. The scale of the concentrated shed vortex forms in the suction surface increases with the increasing of the transverse pressure gradient during the hub-corner separation. When the streamwise pressure gradient increases, the suction surface vortex forms the corner stall. By reasonable design, the two vortexes can cancel out each other. At this time, the loss of cascades is the minimum. Based on the flow mechanism of the corner separation/stall, the trailing gaps are set on three typical turn angle cascades. The results show that the trailing gaps can control the radial development of the suction surface vortex during the stall and improve flow field. The jet cannot blow the suction side boundary layer away during the corner separation, because the gap does not change the static pressure distribution at the root of the cascade. In a word, the trailing edge gaps can not only inhibit the separation in the hub corner but also have the minimum leakage loss at design point. It can be used as an effective and practical compressor design method.

Nowadays, the corner separation, occurring near the corner region formed by the suction surface of blade and end wall, has been an important limitation for the increasing of the aerodynamic loading in the compressor. The previous numerical studies indicate that the Blended Blade and End Wall (BBEW) technology is useful in delaying, or reducing, or even eliminating the corner separation. To further validate the concept, this paper presents combined experimental and numerical investigations on a BBEW cascade and its prototype. Firstly, the NACA65 linear compressor cascade with the turning angle 42 degrees was designed and tested in a low-speed wind tunnel. Then, the cascade with blended blade and end wall design was made and tested in the same wind tunnel. The experimental results show that the design of blended blade and end wall can improve the performance of the cascade when the incidence angle was positive or at the design point, and the total pressure loss coefficient was reduced by 7%–8%. The performance improvement mainly located from 10%–25% span heights. Secondly, based on the experimental data, the numerical study made by our internal code Turbo-CFD shows the difference of the simulation precision of the results, obtained from four different turbulence model after the mesh independence test. The four turbulence model is Spalart-Allmaras model, standard k-ε model, standard k-ω model, and shear stress transport k-ω model. For this case, the SST turbulence model has better performance compared with others. Thirdly, based on the results which were calculated with the turbulence model SST, the effect of the blended blade and end wall design was discussed. The numerical study shows that the design with the blended blade and end wall can have a good effect on the corner flow of the cascade. The strong three-dimensional corner separation, caused by the accumulation of the flow happening at the trail of the suction side was avoided, and the flow losses of the prototype cascade were reduced. Above all, the experiment shows that the design with blended blade and end wall can improve the performance of the cascade. Compared with the experiment data, the SST turbulence model shows the best results of the flow field. Based on the numerical results, the details of the flow field and the effect of the blended blade and end wall design on the corner separation are discussed and analyzed.

The effect of turning angle on the loss generation of Low Pressure (LP) Turbines has been investigated experimentally in a couple of turbine high-speed rigs. Both rigs consisted of a rotor-stator configuration. All the airfoils are high lift and high aspect ratio blades that are characteristic of state of the art LP Turbines. Both rigs are identical with exception of the stator. Therefore, two sets of stators have been manufactured and tested. The aerodynamic shape of both stators has been designed in order to achieve the same spanwise distribution of Cp (Pressure coefficient) over the airfoil surface, each one to its corresponding turning angles. Exit angle in both stators is the same. Therefore the change in turning is obtained by a different inlet angle. The aim of this experiment is to obtain the sensitivity of profile and endwall losses to turning angle by means of a back-to-back comparison between both sets of airfoils. Because the two sets of stators maintain the same pressure coefficient distribution, Reynolds number and Mach number, each one to its corresponding velocity triangles, one can state that the results are only affected by the turning angle. Experimental results are presented and compared in terms of area average, radial pitchwise average distributions and exit plane contours of total pressure losses. CFD simulations for the two sets of stators are also presented and compared with the experimental results.

Following three decades of research in short duration facilities, Purdue University has developed an alternative turbine facility in view of the modern technology in computational fluid mechanics, structural analysis, manufacturing, heating, control and electronics. The proposed turbine facility can perform both short transients and long duration tests, suited for precise heat flux, efficiency and optical measurement techniques to advance turbine aero-thermo-structural engineering. The facility has two different test sections, linear and annular, to service both fundamental and applied research. The linear test section is completely transparent for visible spectra, aimed at TRL 1 and 2. The annular test section was designed with optical access to perform proof of concepts as well as validation of turbine components at the relevant non-dimensional parameters in small engine cores, TRL 3 to 4. The large mass flow (28 kg/s) combined with a minimum hub radius to tip radius of 0.85 allows high spatial resolution. The Reynolds (Re) number extends from 60,000 to 3,000,000, based on the vane outlet flow with an axial chord of 0.06 m and a turning angle of 72 deg. The pressure ratio can be independently adjusted, allowing for testing from low subsonic to Mach 3.2. To ensure that the thermal boundary layer is fully developed the test duration can range from milliseconds to minutes. The manuscript provides a detailed description of the sequential design methodology from zero-dimensional to three-dimensional unsteady analysis as well as of the measurement techniques available in this turbine facility.

Gas turbines are widely used as the marine main power system with its higher power density, react quickly, such as LM2500 and MT30. However, it works under design conditions only during running times of 3% to 10%, and it works under part load during most of the time, leading to low efficiency, and it could not achieve full speed or braking at an instant if sudden emergencies happen. Variable geometry turbines can improve this condition by variable angle nozzle (VAN) technology. And, it could enhance engine braking ability, reduce the fuel consumption under part load, improve the aerodynamic performance of engines, enhance accelerating ability of engines, and implement stalling protection to the power turbine. However, the VAN adjustment needs complicated regulating systems, which makes it difficult to turbine structural design, and leads to increased weight. Besides, there is a performance penalty associated with the vane-end part radial clearance required for the movement of variable vanes. In order to increase the part load efficiency of an intercooled recuperated gas turbine, the power turbine is converted from fixed to variable geometry. And, in order to reduce the losses caused by the radial clearance both of vane ends while vane turning, spherical ends are introduced to keep the clearance constant at all turning angles, and the baseline clearance is 0.77% of blade span. In order to determine the effects of VAN on aerodynamic performance of a variable vane, experimental investigations with a variable geometry turbine annular sector cascade have been conducted under five different turning angles (−6°, −3°, 0°, +5° and +10°) and three Mach numbers (0.3Ma, 0.5Ma and 0.6Ma). The parameter distributions were measured at cascade downstream by a five-hole probe and three-axis auto-traversing system, including outlet flow angle, total pressure loss coefficient, energy loss coefficient. The sector measurement results show that, as the vane turning angle is changed from closed to open, the outlet flow angle are increased under all three test Mach number conditions, which affects the flow mismatching between variable vane and downstream row. And, the total pressure losses is increased with the turning angle changed from design to closed or open, and the total pressure loss increases much more when the vane is closed than when it is open. In addition, vane-end clearances have significantly effects on the flow field. Especially on the hub, the leakage loss is higher, that may be due to the adverse effect of intermediate turbine ducts. Detailed results about these are presented and discussed in the paper.

Flow separations often take place in the junction of blades and endwalls and limit seriously the aerodynamic loading increase of turbomachinery, which are caused mainly by mixing of the boundary layers on blades and endwall surfaces and the transverse secondary flow generated by the pressure difference between the pressure and suction side. Firstly, focusing on a linear diffusion cascade with 42 degrees turning angle, it can be found that the transverse secondary flow can be reduced by inviscid hub and the flow separation is eliminated further through the numerical comparison between the viscous and inviscid hub cases. So the transverse secondary flow is the dominate factor for the flow separation in this cascade. We should try to control the transverse secondary flow to reduce the flow separation. Secondly, based above analysis, the flow separation can be controlled effectively if we can cut off the secondary flow. So nine kinds of streamwise groove schemes are designed and analyzed. It can be seen that the streamwise grooves at the end wall inhibit obviously the transverse secondary flow but the flow structure change is different at different span. There is an optimum combination of width and height of groove, and the height is more important than width. Thirdly, the detailed flow analysis of best scheme with smaller width, moderate height are carried out. It can decrease the separation zone scope at the corner zone, reduce the energy loss coefficient and also reduce the flow loss.

The influence of low to moderate Reynolds number and low to moderate turbulence level on aerodynamic losses is investigated in an incidence tolerant turbine blade cascade for a variable speed power turbine. This work complements midspan heat transfer and blade loading measurements which are acquired in the same cascade at the same conditions. The aerodynamic loss measurements are acquired to quantify the influence of Reynolds number and turbulence level on blade loss buckets over the wide range of incidence angles for the variable speed turbine. Eight discrete incidence angles are investigated ranging from +5.8° to −51.2°. Noting that the design inlet angle of the blade is 34.2° these incidence angles correspond to inlet angles ranging from +40° to −17°. Exit loss surveys, presented in terms of local total pressure loss and secondary velocities have been acquired at four exit chord Reynolds numbers ranging from 50,000 to 568,000. These measurements were acquired at both low (∼0.4%) and moderate (∼4.0%) inlet turbulence intensities. The total pressure losses are also presented in terms of cross passage averaged loss and turning angle. The resulting loss buckets for passage averaged losses are plotted at varied Reynolds numbers and turbulence condition. The exit loss data quantify the impact of Reynolds number and incidence angle on aerodynamic losses. Generally, these data document the substantial deterioration of performance with decreasing Reynolds number.

This paper described a detailed experimental study to explore an internal cooling passage that mimic a “zig-zag” pattern. There are four passages connected by 110° turning angle in a periodic fashion, hence the name. Experiments are performed in a scaled-up test channel with a cross-section of 63.5mm by 25.4mm, corresponding to the aspect ratio of 2.5:1. Compared to the conventional straight internal cooling passages, the zig-zag channel with several turns will generate additional secondary vortices while providing longer flow path that allows coolant to remove much more heat load prior to discharge into the hot mainstream. Surface features, (1) dimples, and (2) protrusions are added to the zig-zag channel to further enhance the heat transfer, while contributed to larger wetted area. The experiment utilizes the well-established transient liquid crystal technique to determine the local heat transfer coefficient distribution of the entire zig-zag channel. Protrusions exhibit higher heat transfer enhancement than that of dimples. However, both designs proved to be inferior compared to the rib-turbulators. Pressure loss in these test channels is approximately twofold higher than that of straight smooth test channel due to the presence of turns; but the pressure loss is lower than the zig-zag channel with rib-turbulators. The result revealed that one advantage of having either protrusions or dimples as these surface elements will resulted in gradual and more uniform increment of heat transfer throughout the entire channel compared to previous test cases.

On the basis of experimental results the new design of a Variable Inlet Guide Vane (VIGV), as can be used for the control and regulation in multishaft compressors, is presented. Main goal of this investigation is a significant increase of the operating range and a reduction of the total pressure loss compared to a currently used basic design. For both designs 2D-cascades were build for detailed measurements in the High-Speed Cascade Wind Tunnel at the Institute of Jet Propulsion at the Universität der Bundeswehr München. The basic design exhibits a symmetric profile with only one segment. In contrast to that the new VIGV design consists of two symmetric vane segments which are arranged pivotable to each other. This provides the advantage of a symmetric profile for a fully opened VIGV associated with a low-loss level. For guidance of the flow, both vane segments can be rotated. Hence, the turning of the flow is split onto two segments. This avoids a huge flow separation on the suction side for high turning angles (Δ β > 30°) which is linked with a strong and abrupt loss increase. Due to the design, the new VIGV exhibits a gap between the two vane segments. Results with opened and sealed gap are presented and discussed. Using a sealing between the segments, a reduction of the profile loss could be detected for all investigated operating conditions. Even without a sealing in the gap, the “low-loss working range” is significantly increased. In addition, it is depicted that the presented results are valid for varying inflow velocities. This broadens the usability of the outcomes. Concluding, it is shown that all aims are achieved. Using the new VIGV design with sealing the low-loss working range can almost be doubled (Δ β > 55°) and the total pressure loss decreases in every working condition compared to the basic design.

Variable geometry turbines are widely used to improve the part-load performance of gas turbine engines. However, there is a performance penalty associated with the vane-end clearance required for the movement of variable vanes. Especially for variable geometry turbines with high casing-endwall angles, greater vane-end clearances are necessary due to annulus slope, and then high endwall leakages would occur, which further deteriorates turbine efficiency. The variable geometry design of the first stage stator vane in a four-stage power turbine featuring very high endwall angles has been carried out by proposed stepped spherical endwall concept. The vane endwalls are spherically shaped so as to maintain constant endwall clearance at all turning angles. And, downstream of the spherical endwall an endwall step is introduced, in order to match the original S-shaped endwall contour and to reduce the leakage loss. Meantime, the rotating shaft is inclined upstream to further match the original endwall contour, and cavity tip design has been used to further reduce the leakage loss. An efficient numerical method has been employed to validate the variable geometry design as mentioned, and the effect of a rotating shaft has been included in the calculations. Then, the four-stage variable geometry power turbine characteristics are evaluated. Results show that the proposed stepped spherical endwall concept can be applied to the variable geometry design of the power turbine featuring very high endwall angles, and compared to the fixed geometry turbine, the efficiency of the new-designed variable geometry power turbine keeps nearly unchanged. Detailed results from this investigation are well presented and discussed in this paper.

Aerodynamic performance of a variable-geometry axial-flow compressor inlet guide vane configuration for a gas turbine unit was determined in a series of annular cascade tests. The variable-geometry vanes used uncambered, symmetrical airfoil sections as the basic blade profile with the rear 70 percent of the vane profile movable as a trailing-edge flap. Vane flap mechanical setting angles of 0 to 50 deg measured from the axial direction were possible, and performance parameters were determined over this range of angles. Turning angles followed a general trend obtained with Carter’s rule for accelerating cascades with the presently measured values tending to be lower than those obtained with Carter’s rule at higher setting angles. For large camber angles (greater than 35 deg) zero-incidence blade element total-pressure loss coefficients for the 50 percent passage location of the flapped vanes tested were higher than those that might have been obtained with a continously cambered vane row of the same solidity and camber.

The aerodynamic coefficients of compressor blade sections in two-dimensional flow can easily and very accurately be determined by use of the well-known Lieblein correlations. Very often the flow across the compressor blade sections is quasi-two-dimensional with the axial velocity density ratio (AVDR) differing from unity. To establish simple correlations for this type of flow as well, the AVDR effect on the aerodynamic coefficients of compressor cascades is theoretically and experimentally investigated. This results in simple but accurate formulas for the calculation of the AVDR effect on the turning angle, the reference minimum-loss inlet angle, and the losses in terms of the wake momentum thickness and the diffusion ratio.

The paper describes theoretical and experimental investigations on the combined effect of axial velocity density ratio (AVDR) and aspect ratio (AR) on compressor cascade performance in incompressible and compressible flow. The results presented demonstrate that it is the aspect ratio that defines the axial velocity distribution through the cascade at a given wall shape and contraction ratio. It is further shown that it is, in turn, the axial velocity distribution that decisively determines the local values of pressure distributions as well as the cascade overall parameters like turning angles and loss coefficients.

Results of an experimental investigation of the aerodynamic performance of several annular prediffuser-combustor systems are presented. Three curved wall, dump prediffusers of different length, area ratio, and turning angle were tested with and without a simulated combustor located downstream of the prediffuser performance was significantly influenced by the presence of the combustor. Pressure recovery and flow losses were determined as a function of prediffuser inlet velocity profile, flow extraction at the prediffuser inlet, axial and radial location of the combustor front end, and distribution of flow in the combustor. Axial location of the combustor was found to be the most significant parameter influencing system performance.

Theoretical evaluation and experimental research have been conducted to verify the performance of steam and gas turbines, including exhaust turbines of superchargers. The simplified channel method and hydroelectrical analog method have been used to calculate blade surface velocity distribution. Based on the “fully developed turbulence” assumption, viscous effects are approximately taken into account by using the boundary layer theory. Theoretical optimum profile loss coefficients are given. Effects of velocity profile on losses are analyzed. Turbine cascades have the characteristics of high solidity, high setting angle and high air turning angle, which facilitate the use of the channel concept. On this basis, K.M. Todd’s “passage convergent gradient,” modified O. Zweifel’s “tangential load coefficient” and other effective criteria have been chosen and cascade data correlated. Some relatively accurate semi-empirical formulas for predicting the aerodynamic performance of cascades are formulated.

A laser-Doppler anemometer was used to measure the three-dimensional velocity field within a typical turbine blade cascade. The blades had a 12.7 cm chord, a turning angle of 104.8°, and a shape conforming to the camber line of a commercial turboexpander. The cascade was operated at a Reynolds number of 1.25×10 5 . Strong secondary velocities, ranging up to 35 percent of the primary flow velocity, were found, resulting from the development of counter-rotating vortices within the blade passages. Large midspan velocity defects in the primary flow were coincident with these high secondary flows. The secondary flow persisted throughout the near wake region.