Karl Engel

Effects of Vortex Generator Application on the Performance of a Compressor Cascade

The performance of a compressor cascade is considerably influenced by secondary flow effects, like the cross flow on the end wall as well as the corner separation between the wall and the vane. An extensive experimental study of vortex generator application in a highly loaded compressor cascade was performed in order to control these effects and enhance the aerodynamic performance. The results of the study will be used in future projects as a basis for parameterization in the design and optimization process for compressors in order to develop novel nonaxisymmetric endwalls as well as for blade modifications. The study includes the investigation of two vortex generator types with different geometrical forms and their application on several positions in the compressor cascade. The investigation includes a detailed description of the secondary flow effects in the compressor cascade, which is based on numerical and experimental results. This gives the basis for a specific approach of influencing the cascade flow by means of vortex generators. Depending on the vortex generator type and position, there is an impact on the end wall cross flow, the development of the horse shoe vortex at the leading edge of the vane, and the extent of the corner separation achieved by improved mixing within the boundary layer. The experiments were carried out on a compressor cascade at a high-speed test facility at DLR in Berlin at minimum loss (design point) and off-design of the cascade at Reynolds numbers up to Re = 0.6 × 106 (based on 40-mm chord) and Mach numbers up to M = 0.7. At the cascade design point, the total pressure losses could be reduced by up to 9% with the vortex generator configuration, whereas the static pressure rise was nearly unaffected. Furthermore, the cascade deflection could be influenced considerably by vortex generators and also an enhancement of the cascade stall range could be achieved. All these results will be presented and discussed with respect to secondary flow mechanisms. Finally, the general application of vortex generators in axial compressors will be discussed.

Experimental Investigation of Flow Control in Compressor Cascades

A large part of the total pressure losses in a compressor stage is caused by secondary flow effects like the separation between the wall and the vane i.e., a corner separation. An experimental and numerical investigation in a highly loaded compressor cascade was performed to understand the fluid mechanic mechanism of this corner separation in order to control it by using vortex generators. The experiments were carried out with a compressor cascade at a high-speed test facility at DLR in Berlin. The cascade consisted of five vanes and their profiles represent the cut at 10% of span distance from the hub of the stator vanes of a single stage axial compressor. The experiments were accomplished at Reynolds numbers up to Re = 0.6 × 106 (based on 40 mm chord) and Mach numbers up to M = 0.7. To measure the total pressure losses of the cascade (caused by the corner separation) a wake rake was used. It consisted of 26 pitot probes to measure the total pressure distribution of the outflow and 4 Conrad probes to determine the outflow angles. To detect the separation area on the vane, a flow visualisation technique was used. In addition to the experiments, numerical computations were carried out with the URANS TRACE, which has been developed at DLR for the simulation of steady and unsteady turbomachinery flow. The computations were performed with identical geometrical conditions as in the experiments, including the measured inflow boundary layer conditions at the side walls. The experiments were performed with the aim of controlling the corner separation. In this case, vortex generators as a passive flow control device were used. The vortex generators were attached at the surface of the suction side of the vanes. The flow control device is producing a strong vortex, which enhances the mixing between the main flow and the retarded boundary layer at the side wall. Thus, the corner separation is reduced on the vanes. The experiments were carried out at the peak efficiency (design point) of the cascade in order to optimize the design of the vortex generators for an application in turbomachines.Copyright

Influence of Blade Fillets on the Performance of a 15 stage Gas Turbine Compressor

In the modern process of the aerodynamic design of multistage compressors and turbines for jet engines as well as for stationary gas turbines, 3D-CFD plays a key role. Before building the first test rig several designs have been investigated using numerical simulations. To understand the characteristics of the individual components it is necessary to simulate their behavior in a multistage simulation and investigate for example, the single stage maps of the compressor in order to understand how the load is divided between the different parts of the compressor during throttling. Increasing computing resources allow ever more details to be incorporated in a 3D simulation. In former times only single blade rows were investigated with a high resolution of the boundary layers, whereas in multistage configurations wall functions were state of the art. Today we are able to apply Low Reynolds resolution even for multistage configurations, so the designer is required to include more and more geometrical details into the simulation. One important such feature is the fillets of rotor and stator blades. Fillets reduce the flow deflection at the endwalls and therefore the loading of the downstream blade rows. This effect is accumulated in a multistage simulation. In this paper a 15-stage compressor with additional inlet and outlet guide vane designed for a stationary gas turbine was investigated with a modern CFD tool by using a real gas approach for two speedlines. Two simulations were done: first a clean configuration with tip and hub clearances but without blade fillets; in the second simulation all rotor blades and the cantilevered stator blades were additionally modeled with fillets. The comparison of the overall global values with measurement data shows a better performance of the simulation with fillets, especially by throttling the compressor. A deeper look into the compressor shows different loads for a considerable number of single stages. The analysis of the steady multistage simulations shows that the numerical stability is reached in different regions of the machine.Copyright

Loss Reduction in Compressor Cascades by Means of Passive Flow Control

Secondary flow effects like the corner stall between the wall and the vane in a compressor stage are responsible for a large part of total pressure losses. An extensive experimental study of flow control in a highly loaded compressor cascade was performed in order to decrease the separation and reduce the losses by means of vortex generators. The vortex generators were attached at the surface of the cascade side walls. These flow control devices produce strong vortices, which enhance the mixing between the main flow and the decelerated boundary layer at the side wall. Thus, the corner flow separation and the total pressure losses could be reduced. The experiments were carried out with a compressor cascade at a high-speed test facility at the DLR in Berlin at minimum loss (design point) and off-design of the cascade at Reynolds numbers up to Re = 0.6 × 106 (based on 40 mm chord) and Mach numbers up to M = 0.7. The cascade consisted of five vanes. The blade profiles are comparable to the hub section of the stator vanes used in the transonic compressor test rig running at Technische Universitat Darmstadt. In the range between −2° and +4° angle of incidence the total pressure losses of the cascade could be reduced up to 4.6% by means of vortex generators, whereas the static pressure rise was not influenced. Based on the results of the cascade measurements, the vortex generators were applied in front of the stator row of the single stage axial compressor at Technische Universitat Darmstadt. A numerical simulation of the compressor flow provided an indication for the adjustment of the vortex generators at the hub and casing. In the experiments the pressure rise and the efficiency of the axial compressor was measured and it could be shown that vortex generators partially improve the efficiency.Copyright

Clocking effects in a 1,5-stage axial turbine : boundary layer behaviour at midspan

This paper presents an experimental and numerical investigation on the influence of clocking on the boundary layer behaviour of the second stator in a 1.5-stage axial low pressure turbine. Surface mounted hot-film sensors were used to measure the quasi shear stress on the second stator and static pressure tappings to obtain the pressure distribution. All experiments were carried out at midspan for different clocking positions. The supporting numerical calculations were conducted with a two-dimensional Navier-Stokes solver using a finite volume discretization scheme and the v′2 f turbulence model.Copyright

CFD Simulations of the TP400 IPC With Enhanced Casing Treatment in Off-Design Operating Conditions

The TP400 intermediate pressure compressor (see Figure 1) is characterized by its extremely wide aerodynamic operating range with strong requirements concerning efficiency and surge margin. Both goals could have been achieved by the proper introduction of variable stator vanes. However, the resulting weight penalty due to the necessary control and actuator system is not accepted — thus this conventional design is rejected and a sophisticated Casing Treatment developed by MTU is introduced. While the underlying multipoint design process is in general expensive and complex the chosen Casing Treatment design (enhanced axial skewed slots [17]) requires the introduction of time accurate 3D CFD simulations in the standard design chain. This ambitious goal leads to the demand for enhanced 3D aerodynamic design tool capabilities like accurate flow prediction in fully turbulent and transitional flow regimes due to different operating conditions as well as the resolution of different geometry features outside the main flow path. In the present paper the effect of different numerical resolution of the “real” geometry as well as the “real” behavior of the flow e.g. steady simulation versus time accurate simulations is discussed. The differences are analyzed and compared to rig-measurements.Copyright

Unsteady Rotor Hub Passage Vortex Behavior in the Presence of Purge Flow in an Axial Low Pressure Turbine

The paper presents an experimental and computational study of the unsteady behavior of the rotor hub passage vortex in an axial low-pressure turbine. Different flow structures are identified as having an effect on the size, strength, shape, position, and the unsteady behavior of the rotor hub passage vortex. The aim of the presented study is to analyze and quantify the sensitivities of the different flow structures and to investigate their combined effects on the rotor hub passage vortex. Particular attention is paid to the effect of the rim seal purge flow and of the unsteady blade row interaction. The rotor under investigation has nonaxisymmetric end walls on both hub and shroud and is tested at three different rim seal purge flow injection rates. The rotor has separated pressure sides at the operating point under investigation. The nondimensional parameters of the tested turbine match real engine conditions. The 2-sensor fast response aerodynamic probe (FRAP) technique and the fast response entropy probe (FENT) systems developed by ETH Zurich are used in this experimental campaign. Time-resolved measurements of the unsteady pressure, temperature and entropy fields between the rotor and stator blade rows are taken and analyzed. Furthermore, the results of URANS simulations are compared to the measurements and the computations are also used to detail the flow field. The experimental results show a 30% increase of the maximum unsteadiness and a 4% increase of the loss in the hub passage vortex per percent of injected rim seal cooling flow. Compared to a free stream particle, the rim seal purge flow was found to do 60% less work on the rotor.

Low Pressure Turbine Secondary Vortices: Reynolds Lapse

A two-stage turbine is tested in a cooperation between the Institute of Aircraft Propulsion Systems (ILA) and MTU Aero Engines GmbH (MTU). The experimental results taken in the Altitude Test Facility (ATF) are used to assess the impact of cavity flow and leakage on vortex structures. The analysis focuses on a range of small Reynolds numbers, from as low as 35,000 up to 88,000. The five hole probe area traverse data is compared to steady multistage CFD predictions behind the second vane. The numerical model compares computations without and with cavities modeled. The simulation with cavities is superior to the approach without cavities. The vortex induced blockage is found to be inversely proportional to the Reynolds number. The circulation of the vortices is dependent on the Reynolds number showing a reversing trend to the smallest Reynolds numbers. The steady numerical model as of yet is unsuitable to predict these trends. A first unsteady simulation suggests major improvements.

Unsteady Wake-Blade Interaction: A Correlation Between Surface Pressure Fluctuations and Loss Generation

This paper presents loss mechanisms associated with unsteady flows in axial turbines due to the interaction of wakes with downstream blade-rows. To evaluate the effect of wake induced loss mechanisms, extensive 3D-URANS computations of an 1.5-stage low pressure turbine rig have been performed. The investigation of the flow field shows how the wake influences well-known effects like the negative-jet effect and resulting flow structures. A detailed view on the impact of these time-dependent flow structures on the fluctuating pressure distribution on the blade is presented by evaluation in both, the time and frequency domains. The effects causing the fluctuations are identified in detail and attributed to their corresponding sources. Unsteady pressure induced by potential effects is distinguished from unsteadiness induced by wake passage. Regarding the latter, it is shown that the amplitude of the pressure fluctuations correlates with the instantaneous state of the wake. In order to describe this behavior, an analytical model is proposed and referred to as the conservation of wake circulation. The evaluation of the entropy generation caused by wake deformation allows the distinction of different loss mechanisms. Finally, an empirical correlation between the amplitude of pressure fluctuation and loss generation is shown.Copyright

Modeling and Analysis of Main Flow-Shroud Leakage Flow Interaction in LP Turbines

Endwall losses significantly contribute to the overall losses in modern turbomachinery, especially when aerodynamic load and pressure ratios are increased. In turbines with shrouded airfoils a large portion of these losses are generated by the leakage flow across the shroud clearance. For the design of modern jet engine turbines it becomes increasingly important to include the impact of shroud leakage flows in the aerodynamic design. There are two main aspects connected to this issue. The first aspect is to optimize the cavity flow and its interaction with the main flow. The second aspect is to perform the airfoil design with boundary conditions, which include the shroud leakage flow effects. In comparison to the simplified approach of neglecting the real endwall geometry and leakage flow this should enable the designer to produce improved airfoils for the entire span. In order to address the second aspect of supporting the airfoil design with improved shroud leakage consideration within the airfoil design process, an efficient procedure for modeling the shroud leakage flow has been implemented into the design Navier-Stokes code. The intention is to model the major leakage flow phenomena without the necessity of pre-defining all details of the shroud geometry. In the paper the results of this model are compared to conventional computations, computations with mesh-resolved cavities and experimental data. The differences are discussed and the impact of certain configuration aspects are analyzed.Copyright