Search

Numerical simulations of particle-laden flows have received growing attention in the last decade, due to the broad spectrum of industrial applications in which discrete phases prediction is of interest. Among these, ingestion of particles by turbomachinery is an area that is seeing vivid research and studies. The most common technique to tackle this kind of problem is the Eulerian-Lagrangian method, in which individual particles are tracked inside the domain. At the same time, in multi-stage turbomachinery simulations interfaces are needed to couple the flow solution in adjacent domains in relative motion. In this work, an open-source extension for Lagrangian simulations in multistage rotating machines is presented in the foam-extend environment. Firstly, a thorough discussion of the implementation is presented, with particular emphasis on particle passage through General Grid Interfaces (GGI) and mixing planes. Moreover, a mass-conservative particle redistribution technique is described, as such a property is requested at interfaces between Multiple Reference Frame (MRF). The peculiarities of the algorithm are then shown on a relevant test-case. Eventually, three turbomachinery applications are presented, with growing complexity, to show the capabilities of the numerical code in real-life applications. Simulation results in terms of erosion and impacts on aerodynamic surfaces are also presented as examples of possible parameters of interest in particle-laden flow computations.

Currently, most of the adjoint-based design systems documented in the open literature assume that the fluid behaves as an ideal gas, are restricted to the optimization of a single row of blades, or are not suited to impose geometric constraints. In response to these limitations, this paper presents a gradient-based shape optimization framework for the aerodynamic design of turbomachinery blades operating under nonideal thermodynamic conditions. The proposed design system supports the optimization of multiple blade rows, and it integrates a computer-aided design (CAD)-based parametrization with a Reynolds-averaged Navier–Stokes (RANS) flow solver and its discrete adjoint counterpart. The capabilities of the method were demonstrated by performing the design optimization of a single-stage axial turbine that employs isobutane (R600a) as working fluid. Notably, the aerodynamic optimization respected the minimum thickness constraint at the trailing edge of the stator and rotor blades and reduced the entropy generation within the turbine by 36%, relative to the baseline, which corresponds to a total-to-total isentropic efficiency increase of about 4 percentage points. The analysis of the flow field revealed that the performance improvement was achieved due to the reduction of the wake intensity downstream of the blades and the elimination of a shock-induced separation bubble at the suction side of the stator cascade.

This paper presents the numerical study of the Source Diagnostic Test fan rig of the NASA Glenn (NASA SDT). Large-Eddy Simulations (LES) based on a finite volume approach are performed for the three different Outlet Guide Vane (OGV) geometries (baseline, low-count and low-noise) and three rotational speeds corresponding to approach, cutback and sideline operating conditions respectively. The full stage and nacelle geometries are considered in the numerical simulations, and results are compared to available measurements. The NASA SDT configuration is equipped respectively with 22 fan blades and either 26 of 54 vanes depending on the OGV geometry. The simulation domain could only be reduced to half of the full annulus and would still be a significant cost for the LES. In order to reduce computational cost, an LES with phase-lagged assumption approach is used. This method allows to perform unsteady simulations of multistage turbomachinery configurations including multiple frequency flows with a reduced computational domain composed of one single blade passage for each row. The large data storage required by the phase-lagged approach is handled by a compression method based on a Proper Orthogonal Decomposition replacing the traditional Fourier series decomposition. This compression method improves the signal spectral content especially at high frequency. Based on the numerical simulations, the flow field is described and used to assess the losses generated in the turbofan configuration based on an entropy approach. The results show different flow topologies for the fan depending on the rotational speed with a leading edge shock at high rotational speed. The fan boundary layer contributes strongly to losses with the majority of the losses being generated close to the leading edge for the dissipation due to mean strains and close to the recirculation zone occurring on the suction side for the turbulent kinetic energy production.

The Covid-19 pandemic has changed the university education, with most teaching moved off campus and students learning online or remote at home, but a cornerstone of undergraduate engineering education has been a big challenge, namely the laboratory classes. As the engineering and education communities continue to adapt to the realities of a global pandemic, it is important to recognize the importance of the laboratory-based courses. In order to address to this situation, an ambitious approach is taken to recreate the laboratory experience entirely online with the help of the digital twins of the fluid mechanics, thermodynamics, and turbomachinery laboratory experiments. Laboratory based undergraduate courses such as EFPLAB1, EFPLAB2 (Energy; Fluid and Process Laboratory 1 & 2) and EFPENG (Energy; Fluid and Process Engineering) are important parts of the “mechanical engineering” and “energy systems engineering” curricula of the Lucerne University of Applied Sciences (HSLU) in Switzerland. Each course mentioned above include six different laboratory experiments about fluid mechanics, thermodynamics, turbomachinery, energy efficiency, and energy systems, including mass- and energy flow balances in energy systems. During the Covid-19 pandemic, it was necessary to adapt to the new environment of remote learning courses and modify the laboratory experiments so that they can be carried out online. The approach was developing digital twins of each laboratory experiment with web applications and providing an environment together with supporting videos and interactive problems so that the laboratory experiments can be carried out remotely. A digital twin is a digital representation of a physical system, e.g., the test rig. It may contain a collection of various digital models with related physical equations and solutions, information related to the operation of the test rig, including 2D or 3D models, process models, sensor data records, and documentation. Ideally, all quantities and attributes that could be measured or observed from the real experiment should be accessible from its digital twin. The digital twin not only reproduces the experimental setup in the laboratory but also helps to improve the knowledge related to the theory and concepts of the laboratory experiments. One major advantage of the digital twin is that the number and range of the parameters, which can be manipulated or varied, is larger in comparison to the actual testing in the laboratory. This paper explains the development of the digital twins (web applications) of the laboratory experiments and provides information about the selected experiments such as potential vortex, linear momentum equation, diffuser flow, radial compressor, fuel cell, and pump test rig with the measurement of pump characteristics. A remote or distance learning has many hurdles, one of the largest being how to teach hands-on laboratory courses outside of an actual laboratory. The experience at the Lucerne University of Applied Sciences showed that teaching online labs using the digital twins of the laboratory experiments can work and the students can take part in remote laboratories that meet the learning outcomes and objectives as well as engage in scientific inquiry from a distance.

The gas turbine engine design involves multi-disciplinary, multi-fidelity iterative design-analysis processes. These highly intertwined processes are nowadays incorporated in automated design frameworks to facilitate high-fidelity, fully coupled, large-scale simulations. The most tedious and time-consuming step in such simulations is the construction of a common geometry database that ensures geometry consistency at every step of the design iteration, is accessible to multi-disciplinary solvers and allows system-level analysis. This paper presents a novel design-intent-driven geometry modelling environment that is based on a top-down feature-based geometry model generation method. In the proposed object-oriented environment, each feature entity possesses a separate identity, denotes an abstract geometry, and carries a set of characteristics. These geometry features are organised in a turbomachinery feature taxonomy. The engine geometry is represented by a tree-like logical structure of geometry features, wherein abstract features outline the engine architecture, while the detailed geometry is defined by lower-level features. This top-down flexible arrangement of feature-tree enables the design intent to be preserved throughout the design process, allows the design to be modified freely and supports the design intent variations to be propagated throughout the geometry automatically. The application of the proposed feature-based geometry modelling environment is demonstrated by generating a whole-engine computational geometry. This geometry modelling environment provides an efficient means of rapidly populating complex turbomachinery assemblies. The generated engine geometry is fully scalable, easily modifiable and is re-usable for generating the geometry models of new engines or their derivatives. This capability also enables fast multi-fidelity simulation and optimisation of various gas turbine systems.

Aircraft operators rely on gas path analysis techniques for monitoring the performance and health of their gas turbine engine assets. This is accomplished by analyzing discernable shifts in measurement parameters acquired from the engine. This paper reviews the founding mathematical principles of gas path analysis, including conventional approaches applied for estimating engine performance deterioration. Considerations for extending the application of gas path analysis techniques to electrified aircraft propulsion (EAP) systems are also discussed, and simulated results from their application to an EAP concept comprised of turbomachinery and electrical system hardware are provided. Results are provided comparing the parameter estimation accuracy offered by taking a whole-system approach toward the problem setup versus that offered by analyzing each subsystem individually. For the latter, the importance of having accurate direct or inferred measurements of external mechanical torque loads placed upon turbomachinery shafts is emphasized.

Labyrinth seals are widely applied in the turbomachinery to control the leakage flow through the clearance between stationary and rotating components. The fluid excitation induced by the labyrinth seal would deteriorate the stability of turbomachinery shaft. Developing an accurate and rapid prediction approach is crucial to the analysis of the fluid excitation rotordynamics of labyrinth seals. The objective of this study is to analyze the applicability of leakage models using Bulk-Flow method and investigate the factors affecting the rotordynamic characteristics of the labyrinth seal. An elliptical orbit for rotor whirling was assumed in the one-control-volume Bulk-Flow model considering an isentropic process to predict the frequency-dependent rotordynamic coefficients of the labyrinth seal. The optimal leakage model was determined by comprehensively analyzing the applicability of seventy-two leakage models. Employing the optimal leakage model in the Bulk-Flow method, the effects of sealing clearance, pressure ratio, preswirl ratio and rotational speed on the rotordynamic characteristics of the labyrinth seal were investigated. The conclusions show that the Bulk-Flow method has an average prediction error of around 10% for the leakage flow rate, cross-coupled stiffness and direct damping when equipped with the optimal leakage model. Increasing preswirl ratio has a significantly destabilizing effect on the rotor stability, while the influence of increasing rotational speed is strongly related to preswirl direction. The effective damping of the labyrinth seal is sensitive to the inlet pressure, but insensitive to the outlet pressure and sealing clearance. The crossover frequency is almost impervious to the inlet pressure, outlet pressure and sealing clearance.

The coupling between different components of a turbomachinery is becoming more widely studied especially by use of computational fluid dynamics. Such simulations are of particular interest especially at the interface between a combustion chamber and a turbine, for which the prediction of the migration of hotspots generated in the chamber is of paramount importance for performance and life-duration issues. Despite this need for fully integrated simulations, typical turbomachinery simulations however often only consider isolated components with either time-averaged constant value, radial profile or least frequently two-dimensional maps imposed at their inlet boundaries preventing any accurate two-way coupling. The objective of this study is to investigate available solutions to perform isolated simulations while taking into account the effect of multicomponent coupling. Investigations presented in the paper focus on the full aero-thermal combustor-turbine interaction research (FACTOR) configuration. The first step of the proposed method is to record conservative variables solved by the large-eddy simulation (LES) code at the interface plane between the chamber and the turbine of a reference simulation. Then, using the spectral proper orthogonal decomposition (SPOD) method, the recorded data is analyzed and can be partially reconstructed using different numbers of frequencies. Using the partial reconstructions, it is then possible to replicate a realistic inlet boundary condition for isolated turbine simulations with both velocity and temperature fluctuations, while reducing the storage cost compared to the initial database. The integrated simulation is then compared to the isolated simulations as well as against simulations making use of averaged quantities with or without synthetic turbulence injection at their inlet. The isolated simulations for which the inlet condition is reconstructed with a large number of frequencies show very good agreement with the fully integrated simulation compared to the typical isolated simulation using average quantities at the inlet. As expected, decreasing the number of frequencies in the reconstructed signal deteriorates the accuracy of the resulting signal compared to the full recorded database. However, isolated simulations with a low number of frequencies still perform better than standard boundary conditions, especially from an aero-thermal point of view.

The technology management methods and the “technology foresight” allow organizations and stakeholders in a particular market/sector to create an advantage out of technological breakthroughs, sustain and expand technological competitiveness, and identify and evaluate new technological options. Several concepts are used depending mainly on the actual status of an evaluated technology. To identify the status of the supercritical carbon dioxide (sCO2) power generation technologies, an extensive “technology exploration” task was performed by creating a technology profile through collecting a database. This allowed for creating technological forecasts “scenario approach” for the sCO2 power cycles. In this article, the sCO2 power technology is explored, evaluated in relation to current commercial competitive power generation technologies, and forecasted to give an insight into future trends of the novel sCO2 power cycle in the future market. The outcome is the analysis of qualitative and quantitative data of the sCO2 technology for short- to long-term forecasts, which could help identify the economic and market value of the sCO2 power cycle. Three possible market scenarios were identified and combined with a survey distributed among experts to assess different market penetration levels of the sCO2 technology.

The adaptive, ℓ2–ω delayed detached eddy simulation model was selected to simulate the flow in the V103 linear compressor cascade. The Reynolds number based on axial chord length is 138, 500. Various inflow turbulent intensities from 0% to 10% were tested to evaluate the performance of the adaptive model. The adaptive model is capable of capturing the laminar boundary layer and the large-scale perturbations inside it. The instability of large-scale disturbances signals the switch to a hybrid simulation of the turbulent boundary layer—thus, the transition front is predicted. In the case of separation-induced transition, the adaptive model, which uses eddy simulation in separated flow, can predict the separation bubble size accurately. Generally, the adaptive ℓ2–ω model can simulate the transitional separated flow in a linear compressor cascade, with a correct response to varying turbulent intensities.

Boundary layer ingestion (BLI) concepts have become a prominent topic in research and development due to their increase in fuel efficiency for aircraft. Virginia Tech has developed the StreamVane™, a secondary flow distortion generator, which can be used to efficiently test BLI and its aeromechanical effects on turbomachinery. To ensure the safety of this system, the complex vanes within StreamVanes must be further analyzed structurally and aerodynamically. In this paper, the induced strain of two common vane shapes at three different operating conditions is computationally determined. Along with these predictions, the aerodynamic damping of the vanes is calculated to predict flutter conditions at the same three operating points. To achieve this, steady computational fluid dynamics (CFD) calculations are done to acquire the aerodynamic pressure loading on the vanes. Finite element analysis (FEA) is performed to obtain the strain and modal response of the StreamVane structure. The mode shapes and steady CFD are used to initialize an unsteady CFD analysis, which acquires the aerodynamic damping results of the vanes. The testcase used for this evaluation was specifically designed to overstep the structural limits of a StreamVane, and the results provide an efficient computational method to observe flutter conditions of stationary systems.

The simulation models used in the design process of modern turbo-machines are becoming increasingly complex. Nevertheless, the steady-state RANS approach is still the mostly used method for CFD computations. However, detailed flow information is more and more required for further improving the performance and extend the operating range. Turbulence modeling has becoming therefore a key issue in this context. The increased computer power already available would enable the use of more sophisticated turbulence models than standard two equation ones, such as SST k-omega and k-epsilon. These, despite their shortcomings, are still predominant in the field, since more advanced models often lead to numerical instabilities in the simulations. The most important shortcomings can be related to either boundary layer effects or mixing process in the channels. In order to improve the predictions considering boundary layer effects, like impingement or large pressure gradients in flow direction, various derivations of four equations models were investigated. Using an additional transport equation for the wall normal Reynolds-stress component and an elliptic equation for near-wall effects, they improve the results for this kind of flows. Considering the accurate prediction of mixing processes, like (1) the interaction of tip-clearance vortices with the main flow or (2) off-design conditions, the focus was oriented to the anisotropy present in the turbulent structures. Standard models are often not sufficient to predict accurately vortices, which can have a huge impact on the performance, since based on the assumption of isotropic turbulence. Accordingly, they tend to dissipate and diffuse the vortices too quickly. Improved models, which take the anisotropic nature of the Reynolds-stresses into account, can help in this context. The models can thereby introduce the additional anisotropy via an explicit algebraic expression, or model directly the transport equations for the Reynolds-stresses. In order to improve the predictions using advanced turbulence models a particularly robust framework based on a pressure-based fully coupled approach was used. The goal of this work is the development and testing of improved models for the application in turbo-machinery. The focus lies thereby on near-wall behavior and mixing / vortex dissipation. The assessment of the models is exemplarily used on the centrifugal compressor open-case Radiver with vaned diffuser.

Probabilistic methods are gaining in importance in aerospace engineering due to their ability to describe the behavior of the system in the presence of input value variance. A frequently employed probabilistic method is the Monte Carlo Simulation (MCS). There, a sample of random representative realizations is evaluated deterministically and their results are afterwards analyzed with statistical methods. Possible statistical results are mean, standard deviation, quantile values and correlation coefficients. Since the sample is generated randomly, the result of a MCS will differ for each repetition. Therefore, it can be regarded as a random variable. Confidence Intervals (CIs) are commonly used to quantify this variance. To gain the true CI, many repetitions of the MCS have to be conducted, which is not desirable due to limitations in time and computational power. Hence, analytical formulations or bootstrapping is used to estimate the CI. In order to reduce the variance of the result of a MCS, sampling techniques with variance reduction properties like Latin Hypercube Sampling (LHS) are commonly used. But the known methods to determine the CI do not consider this variance reduction and tend to overestimate it instead. Furthermore, it is difficult to predict the change of the CI size with increasing size of the sample. In the present work, new methods to calculate the CI are introduced. They allow a more precise CI estimation when LHS is used for a MCS. For this purpose, the system is approximated by means of a meta model. The distribution of the result value is now approximated by repeating the MCS many times. The time consuming deterministic calculations of a MCS are thus replaced with an evaluation on the meta model. These so called virtual MCS can therefore be performed in a short amount of time. The estimated distribution of the result value can be used to estimate the CI. It is, however, not sufficient to use only the meta model. The error ε, defined as the difference between the true value y and the approximated value y, must be considered as well. The generated meta model can also be used to predict the size of the CI at different sample sizes. The suggested methods were applied to two test cases. The first test case examines a structural mechanics application of a bending beam, which features low computational cost. This allows to show that the predicted sizes of the CI are sufficiently precise. The second test case covers the aerodynamic application. Therefore, an aerodynamic Computational Fluid Dynamics (CFD) analysis accounting for geometrical variations of NASA’s Rotor 37 is conducted. For this, the blade is parametrized with the in-house tool Blade2Parameter. For different sample sizes, blades are generated using this parametrization. Their geometrical variance is based on experience values. CFD calculations for these blades are performed with the commercial software NUMECA. Afterwards, the CIs for result values of interest like mechanical efficiency are evaluated with the presented methods. The suggested methods predict a narrower and thus less conservative CI.

Since the 1960s the idea of using supercritical dioxide (S-CO2) as the working fluid in a Brayton power-cycle has been entertained [1]. But due to technical limitations of the time, the idea did not progress forward much. Presently, due to the availability of more knowhow, better technological platform, and advanced analysis tools, many believe it is time to revisit the idea of using supercritical dioxide (s-CO2) as the working fluid for power generation. Among various working fluids, s-CO2 has several significant advantages over other fluids. Primarily the attractive qualities of s-CO2 are high efficiency, much smaller turbomachinery size and plant footprint (and therefore lower capital cost), and the potential for full carbon capture. However, the realization of these benefits will depend on overcoming several technical, engineering and materials science challenges. Even though theoretically, the concept is highly attractive and promising, there is yet a major hurdle to be passed, namely the designing, developing, and testing of a reasonable size (10MWe or higher) prototype of a s-CO2 Brayton-cycle-based power gas turbine. This paper, Part II, in two parts reviews the s-CO2 Brayton cycle technologies for power generation and critically assesses the recent challenges and development status. This present paper (Part-II) in two parts focuses on the turbomachinery components (specifically the compressor) design and challenges.

This paper draws motivation from the fact that engineering optimizations were mostly carried out from scratch. In contrast, however, humans routinely take advantage of the knowledge from past experiences whenever a new task is met. Such a transfer learning process by leveraging knowledge from already completed tasks can be promising to significantly improve the performance of current state-of-the-art algorithms, particularly in solving expensive black-box problems. In light of the above, we propose a Cokriging based transfer optimization framework for the design of turbomachinery cascades, which is demonstrated by optimization to re-design the first-stage vane of GEE3. Specifically, when building Cokriging surrogate in such a transfer optimization context, the samples from already completed tasks are treated as low-fidelity (LF) data. The acquisition function of expected improvement is adopted to guide the search for high-fidelity (HF) data. In order to make full use of the “past experiences”, one of our efforts was drawn to designing selection strategies of initial HF samples. In addition, as the “past experiences” may do harm to the optimization of the target task, the correlation coefficients between source and target tasks in each optimization process were calculated to avoid “negative transfer”. The test results show that, by learning from the past, the transfer optimization framework can reduce the computational cost by much as 50%. More importantly, our proposed transfer learning strategy can effectively avoid “negative transfer” and thus always achieve better solutions than the compared state-of-the-art algorithms.

sCO2 power cycles offer improved cycle efficiencies compared with traditional steam Rankine cycles. However, the turbomachinery required to support such a cycle does not exist at a commercial scale and requires development. This paper describes a new 10 MWe scale sCO2 turbine was developed and demonstrated in an sCO2 closed-loop recompression Brayton cycle. Since this turbine was developed for Concentrating Solar Power (CSP) applications, a target inlet temperature of over 700°C was chosen using funding from the US DOE SunShot initiative and industry partners. However, it can be applied to traditional heat sources such as natural gas, coal, and nuclear power. Traditional Rankine steam cycle thermal efficiencies are typically in the 35–40% range, but can be as high as 45% for advanced ultra-supercritical steam cycles. The sCO2 cycle can approach 50% thermal efficiency using externally fired heat sources. Furthermore, this cycle is also well suited for bottoming cycle waste heat recovery applications, which typically operate at lower temperatures. The high-power density and lower thermal mass of the sCO2 cycle results in compact, high-efficiency power blocks that can respond quickly to transient environmental changes and transient operation, a particular advantage for solar, waste heat, and ship-board applications. The power density of the turbine is significantly greater than traditional steam turbines and is comparable to liquid rocket engine turbo pumps. This paper describes the design and construction of the turbine and provides additional testing of the 10 MWe turbine in a 1 MWe test facility including a description of rotordynamics, thermal management, rotor aero and mechanical design, shaft-end and casing seals, bearings, and couplings. Test data for the turbine is included, as it achieves its operational goal of 715°C, 250 bara, and 27,000 rpm.

A neural network based method is developed that can learn the underlying physics of hydraulic turbocharger (a radial pump coupled with a radial turbine) from a set of sparse experimental data and can predict the performance of a new turbocharger design for any given set of previously unseen operating conditions and geometric parameters. The novelty of the algorithm is that it learns the underlying physical mechanisms from a very sparse data spanning a broad range of flow rates and geometrical size brackets and uses these deeper common features recognized through a “mass-learning process” to predict the full performance curves for any given single geometry. The deep learning algorithm is able to accurately predict the key performance parameters like total efficiency of the turbocharger, its operating speed, pressure rise provided by the radial pump of the turbocharger and the shaft power produced by the radial turbine of the turbocharger for any given input combination of pump and turbine flow rates, differential pressure across the turbine and a limited set of geometrical parameters of pump and turbine impellers and volutes. Lastly, a novel method for fast inverse design of turbomachinery using a physics trained neural network and a constrained optimization algorithms is developed. The algorithm uses Nelder-Mead and Interior Point methods to find the global minimum of turbocharger design objective function in multi-dimensional space. The newly developed method is found to be very efficient in optimizing turbomachinery design problems with both equality and inequality constraints. The inverse design algorithm is able to successfully recommend an optimal combination of geometrical parameters like pump blade exit angle, pump impeller diameter, blade width, eye diameter, turbine nozzle diameter and rotational speed for a given target efficiency and head rise requirements. The preliminary results from this study indicate that it has a great potential to minimize the need for expensive 3D CFD based methods for the design of turbomachinery.

Through-flow methods are widely used in turbomachinery design and performance prediction due to its fast calculation. In the past, circumferentially-averaging through-flow methods tended to ignore circumferential fluctuation source term (CFST) in the governing equations, leading to an inaccurate result when blades showed a strong three-dimensional feature. To solve this problem, CFST needs to be properly modeled. In this paper, three-dimensional numerical simulation is conducted to find the area where CFST dominates. The influence of inlet Mach number, incidence, camber angles and sweep angle on the CFST in this area is analyzed. A model for CFST is constructed. Three-dimensional verification of this model suggests that within the common scope of use of aerospace compressor designing, this model shows a good accuracy.

The dry gas seal is a promising sealing technology to control the leakage flow through the clearance between the stationary and rotational components of Supercritical Carbon Dioxide (SCO2) turbomachinery. The dry gas seal is firstly designed for the SCO2 compressor shaft end of the GE’s 450MWe Brayton cycle power generation system. Then the effects of the spiral angle and gas film thickness on the designed dry gas seal performance are numerically investigated using the three-dimensional Reynolds-Averaged Navier-Stokes (RANS) and SST turbulence model. The accuracy of the numerical method is validated by comparison of the previous research data done by Gabriel et al. with air as the working fluid. The Current study analyzed the sealing performance parameters of the designed dry gas seal for SCO2 compressor shaft end at five gas film thicknesses and four spiral angles. These parameters include: opening force, leakage rate, stiffness, and opening force leakage ratio. Also, the impacts of the spiral angle on flow direction in the fluid film are analyzed. The obtained results show that the designed dry gas seal meets the requirement of the leakage flow rate of the SCO2 compressor shaft end. The dry gas seal with a spiral angle of 15° is the best solution due to its low leakage rate and its’ best comprehensive sealing performance. On some occasions where high stability is required, the dry gas seal with a spiral angle of 30° can be selected due to its’ highest film stiffness. The present work provides the reference of the dry gas seal design for the SCO2 compressor shaft end.

Hydrogen is one of the leading options for storing energy from renewables and surplus electricity. Hydrogen is also a major constituent in various streams in the chemical industry and cannot always be used for better purposes than flaring or heat and power generation. Siemens has identified the 24MWe SGT-600 3rd generation DLE gas turbine as a candidate for having a high hydrogen capability. The burners for using hydrogen in the SGT-600 have been developed for and by Additive Manufacturing technology. The advantages of this technology have been integrated into the presented design and therefore allowing: • Rapid prototyping with possibilities for fast turnaround of tests and screening of various concepts • Manufacturing of complex geometries with smart gas passages, very innovative cooling and mixing concepts • Small series and minimum waste with reduced cost • Good repeatability and stability of product quality Burner development was carried out according to “the standard method within the industry”, meaning CFD-analysis, atmospheric single burner combustion testing followed by pressurized single burner combustion testing and finally a full-scale machine test at the SIEMENS Industrial Turbomachinery AB (SIT) test rig facility in Sweden. The rig is used for full scale testing of gas turbines in the power output range from 15MW to 62MW. It allows testing not only with standard natural gas but also gas mixtures with e.g. hydrogen or nitrogen can be run. The test facility has liquid fuel capability. During the burner development process, a project including two SGT-600 running on up to 60 volume % hydrogen was awarded to Siemens. This meant that a very definite target for the development was set and the results of these efforts are presented in this paper. An adapted 3rd gen. DLE burner design proved to be capable of using 100% hydrogen at SGT-600 full load conditions at the single burner high pressure tests giving only 35 ppm NOx@15%O2. This was a major step in the development of a hydrogen burner for the SGT-600. The following full engine test with the same burner type showed the possibility to run with 60 vol-% H2 at 0–100% load while keeping stable combustion and achieving emissions below 25 ppm NOx@15%O2 in the standard operating range of the SGT-600. At lower loads higher hydrogen contents were tested (95 vol-%) but the flow capacity of the fuel system limited the full exploration of hydrogen capability of the SGT-600 3rd gen. DLE gas turbine. The 3rd gen. DLE burner is also used in the 33MWe SGT-700 and the 62MWe SGT-800, which will also benefit from the development of the increased SGT-600 hydrogen capability. The results open the possibility of using H2 rich gas more widely in all gas turbine configurations using 3rd gen. DLE burner.

The ingestion of multi-mineral dusts by gas turbine engines during routine operations is a significant problem for engine manufacturers because of the damage caused to engine components and their protective thermal barrier coatings. A complete understanding of the reactions forming these deposits is limited by a lack of knowledge of compositions of ingested dusts and unknown engine conditions. Test bed engines can be dosed with dusts of known composition under controlled operating conditions, but past engine tests have used standardised test dusts that do not resemble the composition of the background dust in the operating regions. A new evaporiterich test dust was developed and used in a full engine ingestion test, designed to simulate operation in regions with evaporiterich geology, such as Doha or Dubai. Analysis of the engine deposits showed that mineral fractionation was present in the cooler, upstream sections of the engine. In the hotter, downstream sections, deposits contained new, high temperature phases formed by reaction of minerals in the test dust. The mineral assemblages in these deposits are similar to those found from previous analysis of service returns. Segregation of anhydrite from other high temperature phases in a deposit sample taken from a High Pressure Turbine blade suggests a relationship between temperature and sulfur content. This study highlights the potential for manipulating deposit chemistry to mitigate the damage caused in the downstream sections of gas turbine engines. The results of this study also suggest that the concentration of ingested dust in the inlet air may not be a significant contributing factor to deposit chemistry.

The application of power plants based on renewable energy sources is attractive from an ecological viewpoint. Micro Gas Turbine (MGT) combined with solar energy is a highly promising technology for small-scale electric power generations in remote areas. In MGT state-of-the-art development, the necessity of the numerical optimization in turbomachinery components becomes increasingly important due to its direct impact on the MGT cycle performance. The present paper provides the multidisciplinary design optimization (MDO) of a radial turbine and radial compressor for a 40 kW Solar Hybrid Micro Gas Turbine (SHGT) with a 15m diameter parabolic dish concentrator. The objectives of MDO are to maximize the stage efficiency, to minimize the maximum stress and the inertia, and to enhance the operational flexibility. Preliminary design and performance map prediction using one-dimensional (1D) analysis are performed for both turbine and compressor at various speed lines followed by full three-dimensional (3D) Computational Fluid Dynamics (CFD), Finite Element (FE) analyses and 3D parameterization in the MDO simulations. The purpose of 1D analysis is to set the primary parameters for initial geometry such as rotor dimensions, passage areas, diffuser and volute size. The MDO has been performed using fully coupled multi-stream tube (MST), 3D CFD and FE simulations. MST is used for calculating the load on the blade and the flow distribution from hub to shroud and linearized blade-to-blade calculations based on quasi-three dimensional flow. Thereafter, 3D CFD simulations are performed to calculate efficiencies while the structural stresses are simulated by means of FE analyses. In the current studies, Numeca Fine/Turbo is used as a CFD solver and Ansys Mechanical as a FEA solver, together with Axcent™ as an interface to Fine/Design 3D for geometry parameterization. Furthermore, the cycle analysis for SHGT has been performed to evaluate the effect of the new turbomachinery components from the MDO on the SHGT system performance. It is found that using the MST fully coupled with CFD and FE analysis can significantly reduce the computational cost and time on the design and development process.

Stall flutter of long blades influences the operation safety of the large steam turbines in off-design conditions. As angles of attack are typically high, a partial or complete separation of the flow from the blade surface occurs. The prediction of stall flutter of turbine blades is a crucial task in the design and development of modern turbomachinery units and reliable design tools are necessary. In this work, aerodynamic stability of a linear turbine blade cascade is tested experimentally at high angle of attack +15°, Ma = 0.2 and the reduced frequency of 0.38. Controlled flutter testing has been performed in a travelling wave mode approach for the torsion with the motion amplitude of 0.5°. In addition, ANSYS CFX with SST k-ω turbulent model is used for URANS simulations of a full-scale computational domain. A separation bubble formed on suction surface near the leading edge has been found in CFD results for each blade. Excellent agreement between the experimental and numerical results in stability maps has been achieved for this case under investigation. This is encouraging and both experimental and numerical techniques will be tested further.

This work presents a framework for predicting unknown input distributions for turbomachinery applications starting from scarce experimental measurements. The problem is relevant to turbomachinery where important parameters are obtained using indirect measurements. In this paper a scroll compressor is used as example but the suggested framework is completely general and can be used to infer missing data on material composition (carbon fiber properties, laser melted specimens for additive manufacturing etc) or input data (such as the turbine inlet temperature). Scroll compressors are small devices with a very complex geometry that is difficult to measure. Moreover these compressors are highly sensitive to manufacturing errors and clearances. For these reasons we have chosen this example as an ideal candidate to prove the effectiveness of the framework. An input probability distribution for the scroll height is recovered based on a scarce, synthetic data set. The scroll height is used as an example of a missing distribution for a geometric parameter as it has the highest variance and is challenging to measure experimentally. The framework consists of two main building blocks: an equivalence in a probabilistic sense and a Non-Intrusive Polynomial Chaos formulation able to deal with scarce data. The probabilistic equivalence is defined by a Probability Density Function (PDF) matching approach in which the statistical distance between probability distributions is quantified by either the Kolmogorov-Smirnov (KS) distance or the Kullback-Leibler (KL) divergence. By representing the missing inputs with a generalised Polynomial Chaos Expansion (gPCE) the back-calculation problem can be recast as an optimisation problem in which an arbitrary Polynomial Chaos (aPC) formulation was used to propagate the uncertain input distributions through a computational model of the system and generate a probability distribution for the Quantity of Interest (QoI). The framework has been tested with multiple non-Askey scheme distributions to prove the generality of the proposed approach.

This work investigates the unsteady fluctuation of inducer recirculation stemming from the diffuser stall that occurs near the surge condition. Experiments and unsteady numerical simulation were utilized for the investigation. Inducer recirculation is known to occur near the surge occurrence flow rate, where the flow rate has a positive slope of the performance curve and the recirculation extends to the upstream of the impeller inlet when decreasing the flow rate more. However, few papers have investigated the unsteady phenomenon of the recirculation, even though the surge is what causes it. Clarifying the recirculation phenomenon is essential in terms of expanding the operation range to the lower flow rate for centrifugal turbomachinery. This was our motivation for investigating the unsteady oscillation phenomenon of the inducer recirculation. We investigated a single-stage centrifugal blower with the maximum pressure rise ratio of 1.2 and focused on the flow rates near surge occurrence. The blower was equipped with an open type centrifugal impeller, a vane-less diffuser, and a scroll casing. The blower performance and pressure time-history data were obtained by experiments. Unsteady simulations using large eddy simulation (LES) were conducted to investigate the flow field in the blower for each flow rate. The obtained performance curve showed that the positive slope of the pressure rise at the lower flow rate was due to the impeller stall and that the inducer recirculation extending upstream of the suction pipe near the slope of the curve was flat. LES analysis revealed that this inducer recirculation had two typical fluctuation peaks, one at 20% of the rotation frequency and the other at 95%. We also found that the stall cell at the impeller inlet propagated in the circumferential direction and swirled at almost the same frequency as the impeller rotation. In addition, the fluctuation at the diffuser derived from the diffuser rotating stall propagated to the suction pipe.

While turbocharging is a key technology for improving the performance and efficiency of internal combustion engines, the operating behavior of the turbocharger is highly dependent on the rotor temperature distribution as it directly modifies viscosity and clearances of the fluid film bearings. Since a direct experimental identification of the rotor temperature of an automotive turbocharger is not feasible at an acceptable expense, a combination of numerical analysis and experimental identification is applied to investigate its temperature characteristic and level. On the one hand, a numerical conjugate heat transfer (CHT) model of the automotive turbocharger investigated is developed using a commercial CFD-tool and a bidirectional, thermal coupling of the CFD-model with thermohydrodynamic lubrication simulation codes is implemented. On the other hand, experimental investigations of the numerically modeled turbocharger are conducted on a hot gas turbocharger test rig for selected operating points. Here, rotor speeds range from 64.000 to 168.000 rpm. The turbine inlet temperature is set to 600 °C and the lubricant is supplied at a pressure of 300 kPa with 90 °C to ensure practically relevant boundary conditions. Comparisons of measured and numerically predicted local temperatures of the turbocharger components indicate a good agreement between the analyses. The calorimetrically determined frictional power loss of the bearings as well as the floating ring speed are used as additional validation parameters. Evaluation of heat flow of diabatic simulations indicates a high sensitivity of local temperatures to rotor speed and load. A cooling effect of the fluid film bearings is present. Consequently, results confirm the necessity of the diabatic approach to the heat flow analysis of turbocharger rotors.

Pressure Gain Combustion (PGC) is considered a possible solution to increase gas turbine cycle efficiency, due to the lower entropy generation in the combustion process. However, the highly unsteady flow produced by PGC makes it more difficult to extract work from its exhaust gas. Any outlet restriction downstream of PGC, such as turbine blades, affects its flow field and may cause additional thermodynamic losses. The unsteadiness in the form of pressure, temperature and velocity vector fluctuations has a negative impact on the operation of conventional turbines. Therefore, evaluating early turbine design parameters for such applications is of great interest. Additionally, experimental measurements and data acquisition present researchers with challenges that have to do mostly with the high temperature exhaust of PGC and the high frequency of its operation. Numerical simulations can provide important insights into PGC exhaust flow and its interaction with turbine blades. In this paper, a Rotating Detonation Combustor (RDC) and a row of nozzle guide vanes have been modeled based on the data from literature and an available experimental setup at TU Berlin. Five guide vane configurations with different geometrical parameters have been modeled. URANS simulations were done for all guide vane arrangements to investigate the effect of solidity and blade type representing different outlet restrictions on the RDC exhaust flow. Total pressure loss and velocity fluctuation were computed upstream and downstream of the vanes. The results analzed the connection between total pressure loss and the vanes solidity and thickness to chord ratio. It is observed that more than 57% of the upstream velocity angle fluctuation amplitude was damped by the vanes. Furthermore, the area reduction was found to be the significant driving factor for damping the velocity angle fluctuations, whether in the form of solidity or thickness on chord ratio increment. A further study of the flow field details revealed that the vane passages act as convergent divergent nozzles in the unsteady flow field and no compression wave exists upstream. This RDC exhaust flow investigation is an important primary step from a turbomachinery standpoint, which provided details of blade behavior in such an unsteady flow field.

Over the past years, nonlinear frequency-domain methods have become a state-of-the-art technique for the numerical simulation of unsteady flow fields within multistage turbomachinery as they are capable of fully exploiting the given spatial and temporal periodicities, as well as modelling flow nonlinearities in a computationally efficient manner. Despite this success, it still remains a significant challenge to capture nonlinear interaction effects within the context of configurations with multiple fundamental frequencies. If all frequencies are integer multiples of a common fundamental frequency, the interval spanned by the sampling points typically resolves the period of the common base frequency. For configurations in which the common frequency is very low in relation to the frequencies of primary interest, many sampling points are required to resolve the highest harmonic of the common fundamental frequency and the method becomes inefficient. In addition when a problem can no longer be described by harmonic perturbations that are integer multiples of one fundamental frequency, as it may occur in two-shaft configurations or when simulating the nonlinear interaction in the context of forced response or flutter, then the standard discrete Fourier transform is no longer suitable and the basic harmonic balance method requires extension. One possible approach is to use almost periodic Fourier transforms with equidistant or non-equidistant time sampling. However, the definition of suitable sampling points that lead to well-conditioned Fourier transform matrices and small aliasing errors is an intricate issue and far from straightforward. To overcome the issues regarding multi-frequency problems described above, a new harmonic balance approach based on multidimensional Fourier transforms in time is presented. The basic idea of the approach is that, instead of defining common sampling points in a common time period, separate time domains, one for each base frequency, are spanned and the sampling points are computed equidistantly within each base frequency’s period. Since the sampling domain is now extended to a multidimensional time-domain, all time instant combinations covering the whole multidimensional domain are computed as the Cartesian product of the sampling points on the axes. In a similar fashion the frequency-domain is extended to a multidimensional frequency-domain by the Cartesian product of the harmonics of each base frequency, so that every point defined by the Cartesian product is an integer linear combination of the occurring frequencies. In this way the proposed method is capable of fully integrating the nonlinear coupling effects between higher harmonics of different fundamental frequencies by using multidimensional discrete Fourier transforms within the harmonic balance solution procedure. The aim of this paper is to introduce the multidimensional harmonic balance method in detail and demonstrate the capability of the approach to simultaneously capture unsteady disturbances with arbitrary excitation frequencies. Therefore the well established aeroelasticity testcase standard configuration 10 in the presence of an artificial inflow disturbance, that mimics an upstream blade wake, is investigated. The crucial aspect of the proposed testcase is that a small ratio of the frequency of the inflow disturbance and the blades vibration frequency is chosen. To demonstrate the advantages of the newly proposed multidimensional harmonic balance approach, the results are compared to unsteady simulations in the time-domain and to state-of-the-art frequency-domain methods based on one-dimensional discrete Fourier transforms.

One of the issues of handling large CFD datasets and process them to derive important design correlations is the limitation in automating the post-processing of data. Machine learning techniques, developed to process large unlabelled dataset, can play a key role on this subject. In this work an unsupervised approach to isolate different flow features inside a 2D cascade is proposed and validated. The approach relies on machine learning methods and in particular on Exploratory Data Analysis (EDA) and Principal Component Analysis for the pre-processing of the data and on K-means clustering for the post-processing. The K-means algorithm was trained on a Design of Experiments (DoE) of over 140 cases of 2D linear cascade configurations to identify the boundary layer on the profiles and the wake downstream. Validation resulted in a perfect capability of identifying the regions of interest. Then a possible exploitation of this method is presented, to compute pressure losses downstream of the cascade and train an artificial neural network to make a regression able to extend data to all the possible combinations of geometrical and operating parameters of the cascade. The same algorithm was applied to 3D flow cascades of profiles with sinusoidal leading edges to stress its extrapolation capability in case of flow regimes not present in the training DoE.

Turbomachinery blade rows can have non-uniform geometries due to design intent, manufacture errors or wear. When predictions are sought for the effect of such non-uniformities, it is generally the case that whole assembly calculations are needed. A spectral method is used in this paper to approximate the flow fields of the whole assembly but with significantly less computation cost. The method projects the flow perturbations due to the geometry non-uniformity in an assembly in Fourier space, and only one passage is required to compute the flow perturbations corresponding to a certain wave-number of geometry variation. The performance of this method on transonic blade rows is demonstrated on a modern fan assembly. Low engine order and high engine order geometry non-uniformity (e.g. “saw-tooth” pattern) are examined. The non-linear coupling between the flow perturbations and the passage-averaged flow field is also demonstrated. Pressure variations on the blade surface and the potential flow field upstream of the leading edge from the proposed spectral method and the direct whole assembly solutions are compared. Good agreement is observed on both quasi-3D and full 3D cases. A lumped approach to compute deterministic fluxes is also proposed to further reduce the computational cost of the spectral method. The spectral method is formulated in such a way that it can be easily implemented into an existing harmonic flow solver by adding an extra source term, and can be potentially used as an efficient tool for aeromechanical and aeroacoustics design of turbomachinery blade rows.

In the present work the evolution of the boundary layer over a low-pressure turbine blade is studied by means of direct numerical simulations. The set-up of the simulations follows the experiments by [1], aiming to investigate the unsteady flow field induced by the rotor-stator interaction. The free-stream flow is characterized by high level of free-stream turbulence and periodically impinging wakes. As in the experiments, the wakes are shed by moving bars modeling the rotor blades and placed upstream of the turbine blades. To include the presence of the wake without employing an ad-hoc model, we simulate both the moving bars and the stationary blades in their respective frames of reference and the coupling of the two domains is done through appropriate boundary conditions. The presence of the wake mainly affects the development of the boundary layer on the suction side of the blade. In particular, the flow separation in the rear part of the blade is suppressed. Moreover, the presence of the wake introduces alternating regions in the streamwise direction of high- and low-velocity fluctuations inside the boundary layer. These fluctuations are responsible for significant variations of the shear stress. The analysis of the velocity fields allows the characterization of the streaky structures forced in the boundary layer by turbulence carried by upstream wakes. The breakdown events are observed once positive streamwise velocity fluctuations reach the end of the blade. Both the fluctuations induced by the migration of the wake in the blade passage and the presence of the streaks contribute to high values of the disturbance velocity inside the boundary layer with respect to a steady inflow case. The amplification of the boundary layer disturbances associated with different spanwise wavenumbers has been computed. It was found that the migration of the wake in the blade passage stands for the most part of the perturbations with zero spanwise wavenumber. The non-zero wavenumbers are found to be amplified in the rear part of the blade at the boundary between the low and high speed regions associated with the wakes.

Both LES and DDES are conducted in a low-Reynolds number tip leakage flow model. The DDES uses the SST kω model and employs the same grid with the LES, but the turbulence field diverges from the LES result. Referring to the comparison between LES and DDES, a modification of the zonal function in the DDES model is proposed, which enhances the dissipation of the modeled turbulence thus promote the transition to fully LES in the tip region when the mesh is fine enough. It can generate much finer vortex structure than the original model, including the primary streamwise vortex, induced vortices and the vortex fragments after breakdown. The modification fixes the underestimation of the vorticity and pressure drop at the formation stage of the tip leakage vortex, and generates more reasonable turbulence field and energy spectra. The modified model is introduced to a real rotor simulation at engineering Reynolds number. Compared with the original model on both mean flow field and turbulence field, the modified model shows favorable agreements with the measurements. The study also gives a practical example of using the tip leakage flow model in turbulence modeling.

LES computations have limited applications in turbomachinery predictions because of the formidable amount of resources they require. Due to the exponential increase of requirements with Reynolds number, LES is usually limited to elements with moderate flow velocities and to investigate flows characterized by multiple length and time scales that overlaps. It is the case of combustion, aeroacoustics, unstable range of operations such as stalled conditions. In all these cases LES can provide unique insights on thermo-fluid-dynamics. The main drawback is that LES is not only very expensive, but also extremely sensitive to inflow conditions. A number of studies pointed out that slight differences in LES inflow conditions can result in a very different flow development and therefore different prediction of performance, noise and so on. It is so sensitive that comparison of computations with different inflow conditions and same arrangement for other parameters result in completely diverging results. There are two major solutions to this problem. First, is to use a cyclic inflow channel to generate a fully turbulent profile to feed to the main simulation. Second, the use of a synthetic turbulence model to generate analytically an unsteady turbulent profile. The drawback of the first approach is the fact that a fully developed inflow can be un-realistic for most turbomachinery applications. On the other hand, the second approach was proved not to be able to correctly reproduce the statistics of turbulence and therefore the synthetic inflows do not provide a real solution to the problem. In this paper we discuss a novel method to generate LES inflow conditions, based on adversarial machine learning. In particular we trained a generative adversarial network (GAN) to reproduce the inflow conditions of a channel flow. In this way, the generator of the GAN is trained to correctly reproduce an unsteady turbulent profile, while the discriminator is used during the training phase as an adversarial agent for the generator. During the validation of the method both the discriminator and the generator will be used to validate the proposed methodology.

Data-driven tools and techniques have proved their effectiveness in many engineering applications. Machine-learning has gradually become a paradigm to explore innovative designs in turbomachinery. However, industrial Computational Fluid Dynamics (CFD) experts are still reluctant to embed similar approaches in standard practice and very few solutions have been proposed so far. The aim of the work is to prove that standard wall treatments can obtain serious benefits from machine-learning modelling. Turbomachinery flow modeling lives in a constant compromise between accuracy and the computational costs of numerical simulations. One of the key factors of process is defining a proper wall treatment. Many works point out how insufficient resolutions of boundary layers may lead to incorrect predictions of turbomachinery performances. Wall functions are universally exploited to replicate the physics of boundary layers where grid resolution does not suffice. Widespread wall functions were derived by the observation of a few canonical flows, further expressed as a simple polynomial of Reynolds number and turbulent kinetic energy. Despite their popularity, these functions are frequently applied in flows where the ground assumptions cease to be true, such as rotating passages or swirled flows. In these flows, the mathematical formulations of wall functions do not account for the distortion on the boundary layer due to the combined action of centrifugal and Coriolis forces. Here we will derive a wall function for rotating passages, through means of machine-learning. The algorithm is directly implemented in the N-S equations solver. Cross-validation results show that the machine-learnt wall treatment is able to effectively correct the turbulent kinetic energy field near the solid walls, without impairing the accuracy of the RANS turbulence model in any way.

During the last decades there has been a rise of awareness regarding the necessity to increase energy systems efficiency and reduce carbon emissions. These goals could be partially achieved through a greater use of gas turbine - solid oxide fuel cell hybrid systems to generate both electric power and heat. However, this kind of systems are known to be delicate, especially due to the fragility of the cell, which could be permanently damaged if its temperature and pressure levels exceed their operative limits. This could be caused by degradation of a component in the system (e.g. the turbomachinery), but also by some sensor fault which leads to a wrong control action. To be considered commercially competitive, these systems must guarantee high reliability and their maintenance costs must be minimized. Thus, it is necessary to integrate these plants with an automated diagnosis system capable to detect degradation levels of the many components (e.g. turbomachinery and fuel cell stack) in order to plan properly the maintenance operations, and also to recognize a sensor fault. This task can be very challenging due to the high complexity of the system and the interactions between its components. Another difficulty is related to the lack of sensors, which is common on commercial power plants, and makes harder the identification of faults in the system. This paper aims to develop and test Bayesian belief network based diagnosis methods, which can be used to predict the most likely degradation levels of turbine, compressor and fuel cell in a hybrid system on the basis of different sensors measurements. The capability of the diagnosis systems to understand if an abnormal measurement is caused by a component degradation or by a sensor fault is also investigated. The data used both to train and to test the networks is generated from a deterministic model and later modified to consider noise or bias in the sensors. The application of Bayesian belief networks to fuel cell - gas turbine hybrid systems is novel, thus the results obtained from this analysis could be a significant starting point to understand their potential. The diagnosis systems developed for this work provide essential information regarding levels of degradation and presence of faults in gas turbine, fuel cell and sensors in a fuel cell – gas turbine hybrid system. The Bayesian belief networks proved to have a good level of accuracy for all the scenarios considered, regarding both steady state and transient operations. This analysis also suggests that in the future a Bayesian belief network could be integrated with the control system to achieve safer and more efficient operations of these plants.

The presented work investigates the tone noise of three booster stages designed for a high bypass ratio turbofan. Calculations were performed at the approach operational conditions. The frequency domain numerical method of multistage turbomachines tone noise simulation, developed in CIAM (Central Institute of Aviation Motors) and implemented in 3DAS (3 Dimensional Acoustics Solver) in-house solver was used. The aim was to perform once again the validation of the computational method and to obtain better understanding of the mechanisms of tone noise generation in multistage turbomachines. The results of the investigation were compared with the experimental data obtained in the CIAM C-3A acoustic test facility. In general satisfactory correspondence between the numerical calculations and the experiment was observed. This can be treated as a significant argument for the validity of the frequency domain method of the multistage turbomachinery tone noise calculations.

Nonsynchronous vibrations are a difficult problem to address for turbomachines due to the complex nature of the forcing. Such vibrations can be caused by vortex shedding, flow instabilities, stall cells, or flutter. Testing a design with such excitations can be difficult in practice due to the required forcing. This work demonstrates an experimental excitation method using pulsed air jet excitation to create nonsynchronous vibrations in engine hardware rotating at nominal design speeds. Experimental runs were conducted to excite a number of engine orders (EOs). Blade tip timing was used to measure the blade response without interfering with the blade dynamics. The bladed disk was held at a constant rotational speed while the air jets were pulsed at a sweeping frequency to simulate rotating forcing. Computational models of the physical system were constructed using parametric reduced order models that incorporate the effects of rotational speed and small mistuning. The computational model was used in simulations that mimic the experiment; the forcing was swept across the blades while being pulsed. This results in a system response that cannot be captured using traditional harmonic analyses. The computational and experimental datasets were compared through mistuning values, amplitudes, and the nodal diameter (ND) content in the system response.

Non-contracting annular seals, such as helical groove seals, are widely used between the impeller stages in the liquid turbomachinery to reduce the fluid leakage and stabilize the rotor-bearing system. However, previous literature has expounded that the helical groove seals possess the poor sealing property at low rotational speed condition and face the rotor instability problem inducing by negative stiffness and damping, which is undesirable for liquid turbomachinery. In this paper, to obtain the high sealing performance and the reliable rotordynamic capability for full operational conditions of the machine, two novel mixed helical groove seals, which possess a hole-pattern/pocket-damper stator matching with a helically-grooved rotor, were designed and assessed for a multiple-stage high-pressure centrifugal liquid pump. In order to assess the static and rotordynamic characteristics of these two types of mixed helical groove seals, a three-dimensional (3D) steady CFD-based method with the multiple reference frame theory was used to predict the seal leakage and drag power loss. Moreover, a proposed 3D transient CFD-based perturbation method, based on the multi-frequency one-dimensional stator whirling model, the multiple reference frame theory and a mesh deformation technique, was utilized for the predictions of seal rotordynamic characteristics. The accuracy of the numerical methods was demonstrated based on the experiment data of leakage and rotordynamic forces coefficients of published helical groove seals and hole-pattern seal. The leakage and rotordynamic forces coefficients of these two mixed helical groove seals were presented at five rotational speeds (0.5 krpm, 2.0 krpm, 4.0 krpm, 6.0 krpm, 8.0 kpm) with large pressure drop of 25MPa, and compared with three types of conventional helical groove seal (helical grooves on rotor, stator or both), and two types of damper seals (hole-pattern seal, pocket damper seal with smooth rotor). Numerical results show that the mixed groove seals possess generally better sealing capacity than the conventional helical groove seals, especially at low rotational speed conditions. The circumferentially-isolated cavities (hole or pocket) on the stator enhance the “pumping effect” of the helical grooves for mixed helical groove seals, what is more, the helical grooves also strengthen the dissipation of kinetic energy in the isolated cavities, thus the mixed helical groove seal offers less leakage. Although the mixed helical groove seals possess a slightly larger drag power loss, it is acceptable in consideration of reduced leakage for the high-power turbomachinery. The present novel mixed helical groove seals have pronounced stability advantages over the conventional helical groove seal, due to the obvious large positive stiffness and increased damping. The mixed helical groove seal with the hole-pattern stator and the helically-grooved rotor (HPS/GR) possesses the lowest leakage and the largest effective damping, especially for the high rotational speeds. From the viewpoint of sealing capacity and rotor stability, the novel mixed groove seals are better seal concepts for liquid turbomachinery.

With increasing regulation and the push for clean energy, the operation of power plants is becoming increasingly complex. This complexity combined with the need to optimize performance at base load and off-design condition means that predicting power plant performance with computational modeling is more important than ever. However, traditional modeling approaches such as physics-based models do not capture the true performance of power plant critical components. The complexity of factors such as coupling, noise, and off-design operating conditions makes the performance prediction of critical components such as turbomachinery difficult to model. In a complex system, such as a gas turbine power plant, this creates significant disparities between models and actual system performance that limits the detection of abnormal operations. This study compares machine learning tools to predict gas turbine performance over traditional physics-based models. A long short-term memory (LSTM) model, a form of a recurrent neural network, was trained using operational datasets from a 100 kW recuperated gas turbine power system designed for hybrid configuration. The LSTM turbine model was trained to predict shaft speed, outlet pressure, and outlet temperature. The performance of both the machine learning model and a physics-based model were compared against experimental data of the gas turbine system. Results show that the machine learning model has significant advantages in prediction accuracy and precision compared to a traditional physics-based model when fed facility data as an input. This advantage of predicting performance by machine learning models can be used to detect abnormal operations.

The following paper presents a new type of gas lubricated thrust bearing that utilizes additive manufacturing or also known as direct metal laser melting (DMLM) to fabricate the bearing. The motivation for the new bearing concept is derived from the need for highly efficient supercritical carbon dioxide (sCO2) turbomachinery in the mega-watt power range. The paper provides a review of existing gas thrust bearing technology, outlines the need for the new DMLM concept, and discusses proof of concept testing results. The new concept combines hydrostatic pressurization with individual tilting pads that are flexibly mounted using hermetic squeeze film dampers (HSFD) in the bearing-pad support. This paper describes the thrust bearing concept and discusses the final design approach. Proof-of-concept testing in air for a 6.8” (173mm) outer diameter thrust gas bearing was performed; with thrust loading, up to 1,500 lbs (6.67kN) and a thrust runner speed of 10krpm (91 m/s tip speed). The experiments were performed with a bent shaft resulting in thrust runner axial vibration magnitudes of 2.9mils (74microns) p-p and dynamic thrust loads of 270 lbs (1.2kN) p-p. In addition, force deflection characteristics and stiffness coefficients of the bearing system are presented for an inlet hydrostatic pressure of 380psi (2.62MPa). Results at 10krpm show that the pad support architecture was able to sustain high levels of dynamic misalignment equaling 6 times the nominal film clearance while demonstrating a unit load carrying capacity of 55psi (0.34Mpa). Gas-film force deflection tests portrayed nonlinear behavior like a hardening spring, while the bearing pad support stiffness was measured to be linear and independent of gas film thickness.

The present experimental study expands an ongoing effort to characterize the interactions of axial casing grooves (ACGs) with the flow in the tip region of an axial turbomachine. In recent work, we have tested a series of grooves with the same inlet geometry that overlaps with the rotor blade leading edge, but with different exit directions. Two geometries have stood out: The U grooves, which have an outflow in the negative circumferential direction (opposing the blade motion) are the most effective in suppressing stall, achieving as much as 60% reduction in stall flowrate, but cause a 2% decrease in efficiency around the best efficiency point (BEP). In contrast, the S grooves, which have an outflow in the positive circumferential direction, achieve a milder improvement in stall suppression (36%) but do not degrade the performance near BEP. This paper focuses on explaining these trends by measuring the flow in the tip region and within the U and S grooves. The stereo-PIV (SPIV) measurements are performed in the JHU refractive index matched facility, which allows unobstructed observations in the entire machine. Data has been acquired in two meridional planes that intersect with the grooves at different locations, and two radial planes (z, θ), the first coinciding with the blade tip, and the second, with the tip gap. For each plane, data has been acquired at fourteen rotor orientations relative to the grooves to examine the rotor-grooves interactions. At low flow rates, the inflow into both grooves peaks periodically when the blade pressure side (PS) faces the entrance (downstream side) to the grooves. This inflow rolls up into a large vortex that remains and lingers within the groove long after the blade clears the groove. The outflow depends on the shape of the groove. For the S groove, the outflow exits at the upstream end of the groove in the positive circumferential direction, as designed. In contrast, for the U grooves, the fast radially and circumferentially negative outflow peaks at the base of the U. The resulting jet causes substantial periodic variations in the flow angle near the leading edge of the rotor blade. Close to the BEP, the chordwise location of primary blade loading moves downstream, as expected. The inflow into the grooves occurs for a small fraction of the blade passing period, and most of the tip leakage vortex remains in the main flow passage. For the S grooves, the rotor-groove interactions seem to be minimal, with little (but not zero) inflow or outflow at both ends, and minimal changes to the flow angle in the passage. In contrast, for the U groove, the inflow into and outflow from the groove reverse direction (compared to the low flowrate trends), entering at the base of the U, and exiting mostly at its downstream end, especially when the blade is not near. The resulting entrainment of secondary flows from the groove into the passage are likely contributors to the reduced efficiency at BEP for the U grooves.

Conventional gas turbines are a very mature technology and performance improvements are becoming increasingly difficult and costly to achieve. Pressure Gain Combustion has emerged as a promising technology in this respect, due to the higher thermal efficiency of the respective ideal gas turbine cycles. Up to date, the majority of studies in the field has concentrated on the connection of the turbine to various types of pressure gain combustion systems. The current work focusses on the connection of an array of pulsed detonation tubes at the outlet of an axial compressor. An ideal 0-D plenum is assumed between the compressor outlet and the pulsed detonation tubes’ inlet. The tubes are simulated as time-dependent mass sinks in this plenum. A solution of the time-dependent mass conservation equations delivers the pressure variation in this ideal plenum for various combinations of its size, the tube number, the tube frequency, their firing pattern and their closing time. This pressure variation is subsequently used as a boundary condition in a numeric model of a compressor that can resolve very fast changes of its operational condition. The model is based on the solution of the time-dependent 1-D Euler equations with source terms for the blade forces and the work exchange with the fluid along the compressor stages. These source terms are computed from the operational map of the compressor. The influence of the aforementioned five parameters (number of tubes, frequency, closing time, firing pattern and plenum size) on the compressor operational stability and the flow in its stages is studied in detail. The aim is to benchmark the design of a plenum that results in the lowest possible compressor outlet pressure fluctuation without becoming too large. At the same time, the pressure fluctuations and their propagation along the compressor stages are analyzed.

Flutter stability is a dominant design constraint of modern gas and steam turbines. Thus, flutter-tolerant designs are currently explored, where the resulting vibrations remain within acceptable bounds. In particular, friction damping has the potential to yield Limit Cycle Oscillations (LCOs) in the presence of a flutter instability. To predict such LCOs, it is the current practice to model the aerodynamic forces in terms of aerodynamic influence coefficients, derived for some normal modes of the linearized structural model and fixed oscillation frequency. However, this approach neglects that both the nonlinear contact interactions and the aerodynamic stiffness cause a change in the deflection shape and the frequency of the LCO. This, in turn, may have a substantial effect on the aerodynamic damping. The goal of this paper is to assess the technical importance of these neglected interactions. To this end, a state-of-the-art aero-elastic model of a low pressure turbine blade row is considered, undergoing nonlinear frictional contact interactions in the tip shroud interfaces. The LCOs are computed with a fully-coupled harmonic balance method, which iteratively computes the Fourier coefficients of structural deformation and conservative flow variables, as well as the a priori unknown frequency. The coupled algorithm was tested for various combinations of harmonics in both domains and found to provide excellent computational robustness and efficiency. Moreover, a refinement of the conventional energy method is developed and assessed, which accounts for both the nonlinear contact boundary conditions and the linearized aerodynamic influence. It is found that the conventional energy method may not predict a limit cycle oscillation at all while the novel approach presented here can. Furthermore the refined energy method provides deep understanding of the nonlinear aero-elastic interactions.

Counter-rotating turbomachines have the potential to be high efficiency, high power density devices. Comparisons between conventional and counter-rotating turbomachines in the literature make multiple and often contradicting conclusions about their relative performance. By adopting appropriate non-dimensional parameters, based on relative blade speed, the design space of conventional machines can be extended to include those with counter-rotation. This allows engineers familiar with conventional turbomachinery to transfer their experience to counter-rotating machines. By matching appropriate non-dimensional parameters the loss mechanisms directly affected by counter-rotation can be determined. A series of computational studies are performed to investigate the relative performance of conventional and counter-rotating turbines with the same non-dimensional design parameters. Each study targets a specific loss source, highlighting which phenomena are directly due to counter-rotation and which are solely due to blade design. The studies range from two-dimensional blade sections to three-dimensional finite radius stages. It is shown that, at hub-to-tip ratios approaching unity, with matched non-dimensional design parameters, the stage efficiency and work output are identical for both types of machine. However, a counter-rotating turbine in the study is shown to have an efficiency advantage over a conventional machine of up to 0.35 percentage points for a hub-to-tip ratio of 0.65. This is due to differences in absolute velocity producing different spanwise blade designs.

Turbulent-ribbed channels are extensively used in turbomachinery to enhance convective heat transfer in internally cooled components such as turbine blades. One of the key aspects of such a problem is the distribution of the heat transfer coefficient (HTC) in fully developed flows, and many studies have addressed this problem by the use of computational fluid dynamics (CFD). In the present document, large eddy simulation (LES) is performed for a configuration from a test-rig at the Von Karman Institute representing a square channel with periodic square ribs. The whole channel is computed in an attempt to better understand HTC maps in this specific configuration. Resulting mean and unsteady flow features are captured, and predictions are used to further explain the obtained HTC distribution. More specifically turbulent structures are seen to bring cold gas from the main flow to the wall. A statistical analysis of these events using the joint velocity-temperature probability density function (PDF), and quadrant method allows to define four types of events happening at every location of the channel and which can then be linked to the HTC distribution. First, the HTC is very high where the flow impacts the wall with cold temperature whereas it is lower where the hot gas is ejected to the main flow. In an attempt to link the HTC trace on the channel wall with structures in the flow field far-off the wall, the main modes are identified performing power spectral density (PSD) analysis of the velocity along the channel. Dynamic mode decomposition (DMD) of the flow field data is then used to present the spatio-temporal characteristics of two of the identified most dominant modes: a vortex-street mode linked to the first rib and a rib-to-rib mode appearing because of the quasi-periodicity of the configuration. However, DMD analysis of the HTC trace on the wall does not emphasize any dominant mode. This indicates a weak link between the main flow large scale features and the instantaneous and more local HTC distribution.

The balance piston seal in multiple-stage centrifugal compressors and axial turbines sustains the largest pressure drop through the machines and therefore plays an important role in successful full load operation at high rotational speed. This is especially true for power dense turbomachines in supercritical CO2 power cycles that generate or expend higher fluid pressures (above the critical value 7.3 MPa) and density (close to water 1000 kg/m3), because the fluid forces generated by the balance piston seals are directly proportional to the fluid density and the pressure drop across the seal. This paper presents a comprehensive assessment and comparison on the leakage and rotordynamic performance of three types of annular gas seals for application in a 14 MW supercritical CO2 turbine. These three seals represent the main seal types used in high-speed rotating machines at the balance piston location in efforts to limit internal leakage flow and achieve rotordynamic stability, including a labyrinth seal (LABY), a fully partitioned pocket damper seal (FPDS), and a hole-pattern damper seal (HPS). These three seals were designed to have the same sealing clearance and similar axial lengths. To enhance the seal net damping capability at high inlet preswirl condition, a straight swirl brake was also designed and employed at seal entrance for each type seal to reduce the seal inlet preswirl velocity. Numerical results of leakage flow rates, rotordynamic force coefficients, cavity dynamic pressure, and swirl velocity developments were analyzed and compared for three seal designs at high positive inlet preswirl (in the direction of shaft rotation), using a proposed transient computational fluid dynamic (CFD)-based perturbation method based on the multiple-frequency elliptical-orbit rotor whirling model and the mesh deformation technique. To take into account of real gas effect with high accuracy, a table look-up procedure based on the National Institute of Standards and Technology reference fluid properties database was implemented, using an in-house code, for the fluid properties of CO2 in both supercritical and subcritical conditions. Results show that the inlet swirl brake can significantly reduce the preswirl velocity at seal entrance, lowering the effective damping crossover frequency fco (or even fco = 0) to maximize the full operational frequency range of the machines. In stability analysis phase of a MW-scale supercritical CO2 turbine/compressor, the seal stiffness effects on the rotor mode shape must be evaluated carefully, where the seal stiffness is sufficiently large (comparable to the bearing stiffness). From a rotordynamic viewpoint, the HPS seal with entrance swirl brake is a better seal concept for the balance piston seal in supercritical CO2 turbomachinery, which possesses the largest positive effective damping throughout the entire subsynchronous frequency range.

Labyrinth seals are used to prevent and control the mass flow rate between rotating components. Due to thermally and mechanically induced expansions during operation and transient flight maneuvers, a contact, the so-called rubbing process, between rotor and stator cannot be excluded. A large amount of rubbing process data concerning numerical and experimental investigations is available in public literature as well as at the Institute of Thermal Turbomachinery (ITS). The investigations were carried out for different operating conditions, material combinations, and component geometries. In combination with the experiments presented in this paper, the effects of the different variables on load due to rubbing are compared, and discussed with the focus lying on the material combination. The influence of the material on the loads can be identified as detailed as never before. For example, the contact forces in the current experiments are higher due to a higher temperature resistance of Young’s modulus. The analysis will also be based on the rubbing of turbine blades. Design guidelines are derived for labyrinth seals with improved properties regarding tolerance of rub events. Based on the knowledge obtained, guidelines for designing reliable labyrinth seals for future engines are discussed.

The non-uniform disturbance in the circumferential direction is the essential cause of the occurrence of rotating stall in a turbomachinery. In order to study the effect of tip clearance leakage flow on rotating stall, the mixed-flow pump models with different tip clearances were numerically calculated, and then the energy performance curves and internal flow structures were obtained and compared. The results show that the pump efficiency and the internal flow field of the numerical calculation are in good agreement with the experimental results. A saddle region appeared in the energy performance curves of the three tip clearances, and with the decrease of clearance, the head and efficiency of the mixed-flow pump increased to varying degrees, but the critical stall point shifted to the large flowrate, and the stable operating range of the mixed-flow pump decreased, which indicating that the mixed-flow pumps were easier to stall under small tip clearance. Under the deep stall condition, the influence of the leakage flow on the end wall area increases gradually with the decrease of clearance. Under the small clearance, the leakage flow flows away from the suction surface for a certain distance to form a number of tip leakage vortex strips with the mainstream action, overflows the leading edge of the next blade, and then flows downstream into different flow passages, aggravating backflow and secondary flow separation at the blade inlet, which seriously damaged the spatial structure of the inlet flow. This resulted in the earlier occurrence of stall. With the increase of tip clearance, the tip leakage vortex develops along the radial direction to the middle of the flow channel and serious flow separation phenomenon occurs in the downstream channel, which induces the deep stall. Under the 0.8 mm tip clearance, the whole impeller outlet passage is almost blocked by the backflow of the guide vane inlet, and a deep stall was induced.

Challenges in predicting the turbulence in the tip region of turbomachines include anisotropy, inhomogeneity, and non-equilibrium conditions, resulting in poor correlations between the Reynold stresses and the corresponding mean strain rate components. The geometric complexity introduced by casing grooves exacerbates this problem. Taking advantage of a large database collected in the refractive index-matched liquid facility at JHU, this paper examines the evolution of turbulence in the tip region of an axial turbomachine with and without axial casing grooves, and for two flow rates. The semi-circular axial grooves are skewed by 45° in the positive circumferential direction, similar to that described in Müller et al. [1]. Comparison to results obtained for an untreated endwall includes differences in the distributions of turbulent kinetic energy (TKE), Reynolds stresses, anisotropy tensor, and dominant terms in the TKE production rate. The evolution of TKE at high flow rates for blade sections located downstream of the grooves is also investigated. Common features include: with or without casing grooves, the TKE is high near the tip leakage vortex (TLV) center, and in the shear layer connecting it to the blade suction side tip corner. The turbulence is highly anisotropic and inhomogeneous, with the anisotropy tensor demonstrating shifts from one dimensional (1D) to 2D and to 3D structures over small distances. Furthermore, the correlation between the mean strain rate and Reynolds stress tensor components is poor. With the grooves, the flow structure, hence the distribution of Reynolds stresses, becomes much more complex. Turbulence is also high in the corner vortex that develops at the entrance to the grooves and in the flow jetting out of the grooves into the passage. Consistent with trends of production rates of normal Reynolds stress components, the grooves increase the axial and reduce the radial velocity fluctuations compared to the untreated endwall. These findings introduce new insight that might assist the future development of Reynolds stress models suitable for tip flows.