Abstract Recent efforts have been devoted in developing cutting edge methods and technologies to overcome the complications involved in extracting power from the spools in turbofan engines to drive the power systems in aircraft. In a contemporary turbofan engine design, a Dual Drive Booster turbofan (DDBTF) summation gear box is employed to derive power from the low pressure (LP) and high pressure (HP) spools. This paper aims to investigate the scavenging of lubrication oil from the Dual Drive Booster gearbox. It is essential that that the scavenging of oil from the gearbox is efficient to eradicate risks that may arise when oil resides in the gear box for prolonged durations. Longer residence times of oil in the gearbox can lead to rapid oil degradation. Simulations were conducted on a previously optimized geometry and the work in this paper will focus on investigating the effect of different operating conditions on the scavenging performance of the scavenge chamber. The effect of attitude, altitude and the inlet flow rate of oil have been simulated to understand their influence on the oil flow behavior. Emphasis is given on the predicting potential oil churning, recirculation and pooling behaviors in the scavenge chamber that encloses the gear box. Numerical Investigations are carried out using ANSYS Fluent. The Volume of Fluid (VOF) multiphase model is employed to model the multiphase flow arising between air and oil in the system and the effects of turbulence are modelled using the standard k-ϵ model. The computational domain is discretized using a polyhedral grid comprising of 4 million cells which was adopted based on grid independency tests that were conducted prior to the main simulations. Validation against published experimental data for similar flow regimes was also carried out. Results indicate that the scavenging performance is not affected significantly under the various operating conditions and scenarios that were investigated. This is because the effects of the windage outweigh the effects caused by the different operating conditions that are imposed to the scavenge chamber. The windage in the system drives the oil efficiently out from the chamber with the aid of the tangential sump (shown in Figure 4). Oil is distributed in an axially central section of the chamber and the total residence mass of oil is compared and under 0.5 kg for all the cases presented in this paper.

Abstract Ventilation systems are used in gas turbine packages to control the air temperature, to protect electrical instrumentation and auxiliary items installed inside the enclosure and to ensure a proper dilution of potentially dangerous gas leakages. These objectives are reached only if the ventilation flow is uniformly distributed in the whole volume of the package, providing a good air flow quality as prescribed by international codes such as ISO 21789. To evaluate the effectiveness of the ventilation design, numerical computations are performed for several purposes, one of which is the identification of poorly ventilated portions of the enclosure. In fact, it is essential to accurately detect the regions which are less ventilated, since they could be prone to the accumulation of an accidental fuel gas leak. There are different approaches to identify these portions, such as decay regression or inlet source analysis, that require unsteady simulations of the flow field inside the package. The present work discusses the implementation of a new methodology using machine learning and artificial neural networks (ANN) to detect the poorly ventilated regions where a gas cloud can accumulate. The concentration of fuel gas is estimated starting from a steady-state computation without running a more expensive unsteady computation. The entire process is built around an accurate training of the ANN using a proper set of simpler test-cases that have been identified to match the characteristics of the gas turbine enclosure. During the training phase accuracy and overfitting of the ANN were monitored to ensure robustness of the method. The procedure is then applied to a real case scenario and the results are presented in this paper highlighting the main advantages of this approach respect to a conventional use of CFD analysis. Computations of the flow fields are carried out using OpenFOAM with RANS and U-RANS approaches, while the ANN is developed and trained in Python.

Abstract The development of Ultra-High Bypass Ratio (up to 20) engines with the aim of improving the propulsive efficiency, introduces new challenges for the transmission system in terms of heat management and power losses, since the amount of power transferred through the gearbox is greatly increased. In this respect the accurate estimate of losses at the various flow regimes realized during a typical aeroengine mission within a Power Gear Box (PGB) is essential for the correct design and operation of the engine itself. This paper proposes a computational methodology to estimate all fluid-dynamic (load-independent) losses, which become a major source of dissipation at the high rotational speeds typical of aeroengines, developing within an epicyclic PGB. The overall procedure is based on the superposition principle and approaches the three fluid-dynamic losses, namely injection, windage and meshing losses, with different numerical techniques. The simulations of windage effects, which consider the actual PGB geometry including the carrier disk, the lubricant spray-bar and the external casing, are based on steady-state computations. In order to introduce such simplification, a scaling procedure that avoids interference of the stationary and rotating interfaces was implemented following the outcomes of a previous analysis. The computation of meshing losses employs a fully unsteady dynamic mesh approach and considers a portion of the meshing gears only. Both the sun-planet and planet-ring meshing were considered showing that the latter introduces a much lower level of losses. Finally the injection losses are calculated considering the oil jet momentum variation with simplified methods based on 0D modelling. The proposed procedure, based on the superposition principle and applied to a planetary power gear train, is tested against experimental results described in a previous paper focused on a meshing gear pair.

Abstract This study is aimed at understanding the effects of circumferential groove casing treatment on the performance of a low speed contra-rotating fan. Three dimensional, transient simulations are carried out using the open source CFD library Open-FOAM. The numerical results are validated with experiments for the smooth casing, which show a good agreement. Three treated casing configurations are investigated: 1) grooves at the top of the first rotor, 2) grooves at the top of the second rotor, and 3) grooves at the top of both rotors. Two operating points are simulated for all configurations. Flow inside the grooves is highly dominated by the main and blade passage flows. Grooves increase fan performance at near-stall flow conditions by up to 3 % and reduce pressure fluctuations significantly. Furthermore, they have a larger impact on fan performance when placed on top of the first rotor compared to placing them on top of the second rotor.

Abstract Future aircraft design concepts often show a somewhat wing embedded ultra-high bypass ratio engine. The aircraft concept of the Coordinated Research Centre 880 (CRC880) is a single-aisle configuration with engines partly integrated over the aircraft wing. The aircraft is designed to take off and land on regional airfields with low noise and fuel emissions to address the guidelines set by the ACARE. As a result of the engine installation, the inlet induces a non-axisymmetric boundary layer ingestion into the fan stage. In experimental setups, inlet distortion has often been seen as a 60-degree circumferential inlet stagnation pressure distortion. However, the fan stage inlet flow of the prescribed engine installation of the CRC880 differs to a great extent from a 60-degree sector. In this paper, an aerodynamic comparison between a realistic inflow situation and a similar 60-degree inlet distortion for the same ultra-high bypass ratio fan stage is given. The realistic inflow situation is a result of the flow moving over the aircraft wing suction side and entering the nacelle. As non-axisymmetric inlet geometry remains the same for both cases, therefore, only the total pressure boundary condition at nacelle inlet was changed between both cases. Hence, full annulus simulations are required. Both inlet distortion cases are equivalent by matching average 60-degree distortion coefficient. This study points out that the method, by using averaged 60-degree segment values, excludes specific inflow characteristics. For the same averaged 60-degree distortion coefficient, the local distortion of the embedded case is up to four times larger at rotor tip compared to the segmental approach. For constant mass flow, fan pressure ratio and isentropic efficiency drop by more than five and eight percent respectively. At peak efficiency operating condition, the decrease is even more significant with more than nine percent in stage efficiency. For future embedded aircraft engine configurations, the fan designer has to take into account the non-axisymmetric local flow changes. Hence, the 60-degree segment method is not sufficient when investigating experimental boundary layer ingesting fans and therefore, further method developments are necessary.

Abstract Compressor corner stall is a phenomenon difficult to predict with numerical tools but essential to the design of axial compressors. Predictive methods are beneficial early in the design process to understand design and off-design limitations. Prior numerical work using Navier-Stokes computational methods have assessed the prediction capability for corner stall. Reynolds Averaged Navier-Stokes (RANS) simulations using several turbulence models [1] have shown to over-predict the region of corner hub stall where Large Eddy Simulations (LES) and Detached Eddy Simulations (DES) approaches improved the airfoil surface and wake pressure loss prediction [2, 3]. A linear compressor cascade designed and tested at Ecole Centrale de Lyon [3,4] provides a good benchmark for the evaluation of the accuracy of numerical methods for corner stall. This paper presents results obtained with Lattice-Boltzmann Method (LBM) coupled with Very Large-Eddy Simulations (VLES) approach of PowerFLOW and compares them with the experimental and numerical work from Ecole Centrale de Lyon. The ability to achieve equivalent accuracy at a lower computational cost compared to LES scale resolving methods can enable multi-stage design and off-design compressor predictions. A methodical approach is taken by first accurately simulating the upstream flow conditions. Geometric trips are modeled upstream on the endwalls to match both the mean and fluctuating inflow boundary layer conditions. These conditions were then applied to the simulation of the linear compressor cascade. The benchline experimental study includes trips on both the pressure and suction of the airfoil. These trips are also included for the current simulation. The significance of capturing both inflow conditions and including trips on the airfoil are assessed. Detailed comparisons are then made to airfoil loading and downstream losses between experiment and previous RANS and LES simulations. LBM-VLES corner stall results of pitchwise averaged total pressure match LES agreement relative to experimental data at 50 times lower computational cost.

Abstract The paper discusses the migration, the interaction with the blades, and the attenuation of hot streaks generated by combustor burners, during their propagation within the first turbine stage of aero-engines. Experiments and Computational Fluid Dynamic (CFD) simulations were carried out in the framework of the European Project RECORD and on its follow-up. Measurements considering burner-representative temperature perturbations injected upstream of an un-cooled high-pressure gas turbine stage were performed in the high-speed closed-loop test-rig of the Politecnico di Milano (Italy). The hot streaks were injected in streamwise direction at the stage inlet in four different circumferential positions with respect to the stator blade. They feature a 20% over-temperature with respect to the main flow. Detailed temperature measurements as well as unsteady aerodynamic measurements upstream and downstream of the blade rows were performed. Time-accurate CFD simulations of the flow upstream and within the turbine stage were performed with the TRAF code, developed by the University of Florence. Measurements show a relevant attenuation of hot streaks throughout their transport within the stator and the rotor blade rows, highly depending on the injection azimuthal position. The perturbations were observed to lose their spatial coherence, especially in the transport within the rotor, and to undergo severe spanwise migration. Simulations exhibit a good agreement with the experiments on the measurement planes and allow tracking the complex flow phenomena occurring within the blade rows. Finally the aerodynamic and thermal implications of the inlet temperature perturbations are properly highlighted and discussed.

Abstract Modeling of the engine behavior in windmilling conditions is an important engineering objective. The relight capability of the engine is mainly associated with the air mass flow rate that passes through the engine high pressure core in those conditions. This is one of the parameters that drive the combustion chamber volume. Predicting the engine behavior is challenging, especially early in the development process. The pressure losses along the core are distributed between the different stages of the compressors and turbines, which are operated extremely far from their design point. Engine manufacturers must anticipate with sufficient margins to ensure that the specifications are met when the engine is finally qualified in flight. This article focuses on the behavior of compressor cascades in such conditions, corresponding to high negative incidences. A recently designed high pressure compressor is studied in windmilling conditions using three complementary approaches. First, engine tests are used to obtain validated 0D data for two flight conditions. Then, state of the art Lattice-Boltzmann Method (LBM) simulations are carried out to have insight in the flow characteristics inside the compressor. They are compared to the available experimental data. Finally, a 1D model using stage by stage Euler equation for turbomachinery is used. This kind of modeling is of particular interest because it can be used early in the design process. The correlations for losses and deviation angle from the literature are modified to account for the particularity of the flow in those conditions. One shows that the three approaches give consistent results.

Abstract A comprehensive comparison between Implicit Large Eddy Simulations (ILES) and experimental results of a modern highlift low-pressure turbine airfoil has been carried out for an array of Reynolds numbers ( Re ). Experimental data were obtained in a low-speed linear cascade at the Polithecnic University of Madrid using hot-wire anemometry and LDV. The numerical code is fourth order accurate, both in time and space. The spatial discretization of the compressible Navier-Stokes equations is based on a high-order Flux Reconstruction approach while a fourth order Runge-Kutta method is used to march in time the simulations. The losses, pressure coefficient distributions, and boundary layer and wake velocity profiles have been compared for an array of realistic Reynolds numbers. Moreover, boundary layer and wake velocity fluctuations are compared for the first time with experimental results. It is concluded that the accuracy of the numerical results is comparable to that of the experiments, especially for integral quantities such as the losses or exit angle. Turbulent fluctuations in the suction side boundary layer and the wakes are well predicted also. The elapsed time of the is about 140 hours on 40 Graphics Processor Units. The numerical tool is integrated within an industrial design system and reuses pre- and post-processing tools previously developed for another kind of applications. The trend of the losses with the Reynolds number has a sub-critical regime, where the losses scale with Re −1 , and a supercrital regime, where the losses scale with Re −1/2 . This trend can be seen both, in the simulations and the experiments.

Abstract Understanding distortion transfer and generation through fan and compressor blade rows is able to assist in blade design and performance prediction. Using full annulus URANS simulations, the effects of distortion as it passes through the rotor of a transonic fan at five radial locations (10%, 30%, 50%, 70%, and 90% span) are analyzed. The inlet distortion profile is a 90-degree sector with a 15% total pressure deficit. Fourier distortion descriptors are used in this study to quantitatively describe distortion transfer and generation. Results are presented and compared for three operating points (near-stall, design, and choke). These results are used to explain the relationship between inlet total pressure distortion, pressure-induced swirl, total pressure distortion transfer, total temperature distortion generation, and circumferential rotor power variation. It is shown that very large changes in pressure-induced swirl and distortion transfer and generation occur between near-stall and design, but only small changes are seen between design and choke. The greatest changes are shown to be near the tip. Local power variations are shown to correlate with total pressure distortion transfer and total temperature distortion generation.

Abstract The swirl distortion of a StreamVane™ was investigated in the NASA Glenn Research Center W8 test facility. The StreamVane™ was designed and generated by Virginia Tech based on CFD simulations and included a center body at the aerodynamic interface plane. The swirl pattern generated by the distortion was evaluated using a dense grid of 5-hole Pitot probe measurements captured using a rotating array of probes. Good agreement was found between the design intent and the results at 38.5 kg/s mass flow. The StreamVane™ swirl results were compared to clean facility flow at 5 inlet mass flows and found to be consistent. Additionally, the axial location of the StreamVane™ relative to the measurement plane was investigated to determine the impact on downstream total pressure loss generated by the vanes. The intent of this work was to assess the viability of using a StreamVane™ to generate a Type I or Type II distortion into a Boundary Layer Ingesting propulsor to assess its aerodynamic performance and aeromechanic response.

Abstract Multistage radial compressors are major components in plenty of industrial applications. Today, compressor downsizing for CAPEX reduction is of utmost importance. Accordingly, the pressure ratio must be increased, accommodated in a most compact design, accepting nearly no penalties in the performance level achieved so far. In order to investigate the complex flow in stator parts of multistage centrifugal compressors and to increase the confidence level of numerical methods a new test rig is developed and taken into operation. This test rig allows to separate stationary flow effects from time variant effects leading to an in depth insight into the physical flow features. The aim is to investigate the flow in different stator designs in detail for varying flow coefficients. Additive manufacturing techniques are applied to achieve low costs simultaneously with short production time for the tested parts. In this publucation, the measured flow field in the stator parts is discussed. The experimental results comprise velocity and pressure data from five-hole-probes and unsteady velocity data from single-film hot-film-probes. Using unsteady velocity data, turbulent statistics such as turbulence intensity and the turbulent kinetic energy will be determined. Subsequently, the experimental results are compared to numerical results. Therefore, (U)RANS simulations are performed using a commercial CFD-code. The simulation results are evaluated at planes appropriate to the measuring planes in the test rig to guarantee a directly comparable data base.

Abstract In spite of advances in CFD prediction tools, the current design of outlet guide vane (OGV) stages for flow recovery downstream from low pressure turbines (LPT) still has to face significant flow entrance uncertainties. To ensure proper response of modern, high efficiency OGV’s, the sensitivity in the aerodynamic response of the vanes to both different levels of inlet turbulence and off-design incidence must be analyzed. To that end, a systematic experimental investigation of a current design LPT OGV airfoil has been undertaken in a low-speed linear cascade. Wall pressure distributions as well as high-resolution total pressure drop and LDV measurements have been used to determine the flow response. The experimental facility includes different boundary suction strategies for proper control of flow periodicity and endwall effects at significant off-design incidences. In addition, different inlet grids to promote an entrance flow having controlled isotropic background turbulence are included. The experimental flow response of the OGV airfoil is presented for a wide range of Reynolds numbers and different values of the inlet flow incidence and turbulence properties. Both at design and off-design incidences, different flow regimes and performance degradation mechanisms are discussed. In addition, the effect of inlet turbulence at close to design incidence is discussed, with the experimental evidence suggesting that its effect can be described by defining a properly scaled Reynolds number. The ability of CFD simulations based on currently available RANS transition models to describe the flow in high efficiency turbine OGV airfoils is finally explored.

Abstract In marine diesel engines, transonic centrifugal compressors are widely used due to their capabilities to: 1) downsize engines by increasing output power; 2) cause less fuel consumption; 3) enhance the combustion efficiency. Apart from the traditional requirements such as good choke and stall margin, high boosting pressure ratio, and high stage efficiency, it is also necessary to reduce the turbocharger noise as much as possible for a more comfortable working environment. The most effective way is to reduce the compressor noise at the source, i.e. the compressor impeller itself must produce less aerodynamic noise rather than using any silencers. In this paper, we present a redesign work for an existing impeller wheel using the 3D inverse method with an in-house aeroacoustic code. The CFD simulations for the compressor characteristics and internal flow field details, the numerical predictions of aeroacoustic sound emissions, and FEA analysis for the structure integrity have all been attempted to thoroughly assess the performances of baseline and optimised impellers. The computational results found that the new impeller can lead to better performances in all three aspects, which are supported by the experimental measurements conducted for both impellers using the same test configurations. The experimental data confirmed that the inversely redesigned impeller wheel provides a wider compressor operating range, higher efficiency at large rotating speed, and a few dB(A)s lower noise emissions in the upstream radiation direction.

Abstract Two sound power measurement approaches were developped that are easy to install and have the ability to detect the dominant modal content by applying the modern signal processing method, Compressed Sensing. In general Compressed Sensing requires only few measurement positions for an exact reconstruction of sparse acoustic mode fields. For a current study we have chosen two Compressed Sensing algorithms. Each require separate sensor array arrangements and deliver different modal contents, from which the sound power can be derived. Firstly, an Azimuthal Mode Analysis is conducted by applying the Enhanced Orthogonal Matching Pursuit (EOMP) algorithm to a sound pressure measurement vector. The measurements are obtained by using a sensor ring array with optimized positions. In a subsequent step, the sound power is calculated by referring the detected azimuthal mode spectrum to a model describing the energy distribution over the radial mode content. Secondly, using the Block Orthogonal Matching Pursuit (BOMP) algorithm, the radial mode amplitudes are determined directly. This algorithm requires the sensors to be placed at optimized azimuthal and axial positions and reconstructs a set of dominant radial modes that occur in groups. With the objective to verify both methods, the newly designed and optimized arrays in combination with the aforementioned mode reconstruction algorithms are applied to a numerical data set. This data was provided by URANS simulations of a radial compressor set-up, which is an exact replication of an actual test rig located at the RWTH Aachen University. The introduced estimation methods perform well as shown by comparison with an exact and high resolution Radial Mode Analysis Method. In the near future, the presented measurement approaches will be applied in an experimental study performed at the radial compressor test rig.

Abstract This study carries out parametric investigations on aerodynamic loss of various types of LP turbine airfoils characterized with different flow deceleration rates (DR) on their suction surfaces under the realistic flow conditions such as wake inflow and freestream turbulence. The Reynolds number examined in this study ranges from 57,000 to 170,000. As for the freestream turbulence, two levels of the turbulence are used, i.e., about 1.2% and 3.5%. Stagnation pressure distributions downstream of each of the airfoil cascades are measured by use of a Pitot tube, while steady-state and unsteady boundary-layers are measured over the rear part of suction surface and pressure side near the trailing edge using a single hot-wire probe. The measured boundary-layer data are used to estimate the cascade loss along with RANS (Reynolds-Averaged Navier-Stokes) simulations by taking advantage of the momentum-theory based Denton’s method. First, relationships between the cascade loss for each flow condition and DR are examined. The estimated loss values are then compared with the measured cascade loss to check the validity of the loss estimation method, which is a derivative of Denton’s method, under the realistic flow conditions.

Abstract This paper investigates the generation of rotor-alone tones and their contribution to the outflow noise in a transonic centrifugal compressor stage with vaneless diffuser and volute by means of unsteady full-annulus CFD-simulations. The aerodynamic field, as well as the generation and propagation of sound, were simulated simultaneously using the URANS-approach of the solver TRACE and a numerical grid consisting of 170M cells. To assess the accuracy of the predicted fluctuations, the investigation compares the simulated diffuser flow field to measured flow angles and pressure fluctuations obtained from experiments conducted on a large-scale test rig. The analysis explains the different sound generation mechanisms responsible for tonal components in the acoustic spectrum at the compressor outlet, based on the Fourier decomposition of the pressure fluctuations in diffuser and volute. Further, the paper analyzes the modal structure of the simulated sound field at the volute outlet by means of a radial mode analysis and discusses the influence of changing operating conditions on the sound power emitted. The analyses reveal that supersonic flow phenomena occurring at choked operating conditions cause a significant increase in noise emissions. Furthermore, the investigation shows that the sound field at the volute outlet is dominated by few low-order modes, a fact that justifies analysis using methods based on Compressed Sensing in future experimental investigations.

Abstract The goal of this work is to predict the boundary layer transition induced by a separation bubble on the suction side of a turbine blade of a linear turbine cascade using Large Eddy Simulation (LES). The numerical simulations refer to the linear turbine cascade T106D-EIZ tested at the Institute for Jet Propulsion of the Bundeswehr University Munich (Germany). The blade pitch was increased compared to the design point in order to have a higher load and enhance the formation of a separation bubble at the suction side of the blade. Different flow configurations were tested and the transition of the boundary layer was evaluated. For the numerical case, the operating condition with an inlet turbulence below 1% was used. In the first part of this work, the LES setup is discussed. A modified Smagorinsky subgrid-scale model is used to reduce the turbulent viscosity in the region closest to the wall. The computational grid is designed according to the information coming from the Taylor and the Kolmogorov length scales. These parameters are found from RANS k-omega SST simulations. The fifth-order accurate WENO scheme was used for the computation of the cell fluxes. In the second part of the work, a comparison between the results of the LES simulations and of the RANS k–ω SST simulations with the γ–Re θ transition model is done. Integral and statistical parameters of the boundary layer from the simulations with the two models are evaluated and compared. The ability of the LES and the RANS models to predict the boundary layer evolution along the blade profile and the point of separation will be discussed.

Abstract The intended operating point of turbomachinery is subject to numerous kinds of uncertainty. These range from varying ambient conditions, across geometric deviations in a component, to system related loading variability resulting in engine-to-engine variation in component matching. In order to guarantee safe operation at all conditions, it is essential to consider the above uncertainties when designing turbomachinery. In the present work, a probabilistic assessment is performed of the influence of possible operational uncertainties on the aerodynamic performance metrics of an aero-engine multistage high pressure compressor (HPC). To propagate uncertainties, Monte Carlo simulations (MCS) with Latin Hypercube Sampling (LHS) were performed, with both correlated and uncorrelated inputs. Each sample consisted of a steady state computational fluid dynamics (CFD) evaluation of the compressor. The statistical input for the boundary conditions was acquired from a MCS of the engine cycle performance at cruise, accounting for flight-to-flight variations in ambient conditions and engine-to-engine variations in component properties. With the chosen approach, it is possible to quantify the variability in aerodynamic performance of an HPC that is subject to uncertain operating conditions and thus shows the importance of input correlations. Results highlight that deterministically determined performance metrics can differ considerably from the statistical mean, revealing the benefits of a probabilistic assessment. In contrast to performing MCS on the cycle only, a CFD based assessment can also be used to draw conclusions on the aerodynamic mechanisms responsible for changes in efficiency or surge margin.

Abstract The isentropic exponent is one of the most important properties affecting gas dynamics. Nonetheless, its effect on turbine performance is not well known. The paper discusses a series of experimental and computational studies to determine the effect of isentropic exponent on the flow field within a turbine vane. Experiments are performed using a newly modified transient wind tunnel which enables annular cascade testing with a wide range of working fluids and operating conditions. For the present study tests are undertaken using air, CO 2 , R134a and argon, giving a range of isentropic exponent from 1.08–1.67. Measurements include detailed wall static pressures which are compared with computational simulations. Our results show that over the range of isentropic exponents tested here, the loss can vary by between 20%–35%, depending on vane exit Mach number. The results are important for future turbines operating with real-gas effects and/or those where high gas temperatures can lead to variations in isentropic exponent.