Abstract The paper presents an experimental data base on transitional boundary layers developing on a flat plate installed within a variable area opening endwall channel. Measurements have been carried out by means of time-resolved PIV. The overall test matrix spans 3 Reynolds numbers, 4 free-stream turbulence intensity levels and 4 different flow adverse pressure gradients. For each condition, 16000 instantaneous flow fields have been acquired in order to obtain high statistical accuracy. The flow parameters have been varied in order to provide a gradual shift of the mode of transition from a bypass process occurring with mild adverse pressure gradients at high free-stream turbulence, to separated flow transition, occurring with low Reynolds number, low free-stream turbulence intensity and elevated adverse pressure gradient. In order to quantify the influence of the flow parameter variation on the boundary layer transition process, the transition onset and end positions, and the turbulent spot production rate have been evaluated with a wavelet based intermittency detection technique. This post-processing technique is in fact able to identify the vortical structures developing within the boundary layer, the intermittency function is then automatically evaluated for each tested condition counting the number of such structures and defining the cumulative probability function. The by-pass transition mode has the longest transition length that decreases with increasing the Reynolds number. The transition length of the separated flow case is smaller than the by-pass one, and the variation of the flow parameters has a similar impact. Similarly, the dimensionless turbulent spot production rate reduces when the Reynolds number is increasing. The variation of the inlet turbulence intensity has a small influence on this parameter except for the condition at the highest turbulence intensity, that always shows the lowest turbulent spot production rate because a by-pass type transition occurs. This large amount of data has been used to develop new correlations used to predict the spot production rate and the transition length in attached and separated flows.

Abstract Compressor dynamics were studied in a gas turbine – fuel cell hybrid power system having a larger compressor volume than traditionally found in gas turbine systems. This larger compressor volume adversely affects the surge margin of the gas turbine. Industrial acoustic sensors were placed near the compressor to identify when the equipment was getting close to the surge line. Fast Fourier transform (FFT) mathematical analysis was used to obtain spectra representing the probability density across the frequency range (0–5000 Hz). Comparison between FFT spectra for nominal and transient operations revealed that higher amplitude spikes were observed during incipient stall at three different frequencies, 900, 1020, and 1800 Hz. These frequencies were compared to the natural frequencies of the equipment and the frequency for the rotating turbomachinery to identify more general nature of the acoustic signal typical of the onset of compressor surge. The primary goal of this acoustic analysis was to establish an online methodology to monitor compressor stability that can be anticipated and avoided.

Abstract Modern high pressure turbine (HPT) blade design stands out due to its high complexity comprising three-dimensional blade features, multi-passage cooling system (MPCS) and film cooling to allow for progressive thermodynamic process parameters. During the last decade, probabilistic design approaches have become increasingly important in turbomachinery to incorporate uncertainties such as geometric variations caused by manufacturing scatter and deterioration. Within this scope, the first part of this two-part paper introduces parametric models for cooled turbine blades that enable probabilistic FE analysis taking geometric variability into account to aim at sensitivity and robustness evaluation. The statistical database is represented by a population of more than 400 blades whose external geometry is captured by optical measurement techniques and 34 blades that are digitized by computed tomography (CT) to record the internal geometry and the associated variability, respectively. Based on this data, parametric models for airfoil, profiled endwall (PEW), wedge surface (WSF) and MPCS are presented. The parametric airfoil model which is based on traditional profile theory is briefly described. In this regard, a methodology is presented that enables to adapt this airfoil model to a given population of blades by means of Monte-Carlo based optimization. The endwall variability of hub and shroud are parametrized by radial offsets that are applied to the respective median endwall geometry. WSFs are analytically represented by planes. Variations of the MPCS are quantified based on the radial distribution of cooling passage centroids. Thus, an individual MPCS can be replicated by applying adapted displacement functions to the core passage centroids. For each feature that is considered within the present study, the accuracy of the parametric model is discussed with respect to the variability that is present in the investigated blade population and the measurement uncertainty. Within the scope of the second part of this paper (cf. Högner et al. [1]), the parametric models are used for a comprehensive statistical analysis to reveal the parameter correlation structure and probability density functions (PDFs). This is required for the subsequent probabilistic finite element analysis involving real geometry effects.

Abstract A twin-fluid atomizer configuration is simulated by means of the 2D weakly-compressible Smooth Particle Hydrodynamics method, and compared to experiments. The Gas-to-Liquid-Ratio, the momentum flux ratio and the velocity ratio are set constant for different ambient pressures, which leads to different gaseous flow sections. The objectives of this study are to (i) investigate the effect of ambient pressure at constant global parameters, and (ii) to verify the capability of 2D SPH to qualitatively predict the proper disintegration mechanism and to recover the correct evolution of the spray characteristics. The setup consists of an axial liquid jet of water fragmented by a co-flowing high-speed air stream (U g = 80 m/s) in a pressurized atmosphere up to 16 bar. The results are compared to the experiment, and presented in terms of (i) mean velocity profiles, (ii) drop size distributions and (iii) Sauter Mean Diameter of the spray. It is found that there exists an optimal pressure to minimize the mean size of the spray droplets. Finally, two new quantities related to atomization are presented: (i) the breakup activity that quantifies the number of breakup events per time and volume unit and (ii) the fragmentation spectrum of the whole breakup chain, which characterizes the cascade phenomenon in terms of probability. The breakup activity confirms the presence of the optimal pressure and the fragmentation spectrum gives information on the type of breakup, depending on the ambient pressure.

Abstract Fuel injection plays an important role in liquid fueled gas turbine combustion. The strong interdependence of liquid breakup and atomization, turbulent dispersion of these droplets, droplet evaporation, and fuel-air mixing make the spray modeling an extremely challenging task. The physical processes are even more difficult to predict for alternative fuels with different thermophysical properties. In this study, spray flames of unheated and preheated vegetable oil (VO) produced by an air-blast atomizer in a swirl stabilized combustor are investigated experimentally. Phase Doppler particle analyzer (PDPA) is used to measure the instantaneous diameter and axial velocity of droplets at different axial and radial locations in both flames. Experiments are conducted at an equivalence ratio of 0.79 and atomizing air to liquid ratio (ALR) by mass of 2.5 to obtain stable VO flames. Radial profiles of mean axial velocity and Sauter mean diameter are presented to show the effect of fuel preheating. Joint Probability Density Functions (joint PDF) are presented to show the correlation between droplet diameter and axial velocity. Results are analyzed to show that both sprays exhibit self-similar droplet diameter distributions at different axial and radial locations when normalized properly. Thus, the vast amount of PDPA data in the spray can be reduced to simple distribution functions. A method to reconstruct the joint PDF from experimentally determined distribution functions is presented. We envision that the joint PDF approach outlined in this study could be implemented in high-fidelity computational fluid dynamic models to improve spray predictions in future studies.

Abstract It has been shown that the fluctuations of pressure caused by a thermoacoustic instability can affect the mass flow rate of air and atomisation of the liquid fuel inside a gas turbine. Tests with premixed flames have confirmed that the fluctuations of the mass flow rate of air can affect the heat release rate through purely aerodynamic phenomenon but little work has been done to test the sensitivity of the heat release rate to changes in the fuel atomisation process. In this study, a lean-burn combustor geometry is supplied with a fuel spray fluctuation of SMD (Sauter mean diameter) of 20% with respect to the mean value and the heat release rate predicted using Large Eddy Simulation (LES) with combustion predicted using a presumed probability density function (PPDF), flamelet generated manifolds (FGM) method. Previous work has shown that at atmospheric conditions the SMD may fluctuate by up to 16% percent and at low frequencies may be reasonably well predicted by using a correlation based on the instantaneous velocity and mass flow rate of air close to the air-blast atomiser. Analysis of the flow fields highlights a complicated spray, flame and wall interaction as being responsible for this observed fluctuation of heat release rate. The heat release rate predicted by the LES shows a 20% fluctuation which implies that even small fluctuations of SMD will significantly contribute to thermoacoustic instabilities.

Abstract Combustion characteristics of a can combustor with a rotating casing for an innovative micro gas turbine have been modeled. The effects of syngas compositions and the rotating speed on the combustor performance were investigated. The effects of rotation on the combustion performance have been studied previously with methane as the fuel. This work extended the investigation for future application with syngas blended fuels. Two typical compositions of syngas were used namely: H 2 -rich (H 2 :CO=80:20, by volume) and equal molar (H 2 :CO=50:50). The analyses were performed with a computational model, which consists of three-dimension compressible k-ε realizable turbulent flow model and presumed probability density function for combustion process invoking a laminar flamelet assumption generated by detailed chemical kinetics from GRI 3.0. As syngas is substituted for methane at a constant fuel flow rate, the high temperature flame is stabilized along the wall of the combustor liner. With the casing rotating, pattern factor and exit temperature increase, but the lower heating value of syngas causes a power shortage. To make up the power, the fuel flow rate is raised to maintain the thermal load. Consequently, the high temperature flame is pushed downstream due to increased fuel injection velocity. NO x emission decreases as the rotational speed increases in both cases. Pattern factor decreases but exit temperature increases with the increase of roatation speed indicating a higher combustion efficiency. However, there is possible hotspots at exit due to higher pattern factor (PF>0.3) for H 2 -rich and equal molar syngas at lower speed of rotation, which needs to be resolved by improving the cooling strategy.

Abstract This paper focuses on the development of a structural health monitoring system based on guided Lamb waves propagating over the structure and a network of surface acoustic sensors in communication at high frequencies. A time-of-flight (ToF), algorithm and a probabilistic diagnostic imaging and calibration method is developed to detect miniscule material losses or material adhesion as well as the defects like small scale holes and cracks in turbomachinery components like blades, rotors, plates and pipes. Using an advanced ToF algorithm, precise differences in timescales for arrival of symmetric / antisymmetric lamb wave packets are found for all possible combinations of actuator-sensor pairs. This leads to a deterministic mathematical construct for damage localization for various actuator-sensor pairs at focal points. In the probabilistic diagnostic imaging (PDI) method, field value is assigned based on fusion of wave signals rendered by various actuator-sensor paths to indicate the probability of the presence of a damage at a particular location on the structure. Correlation coefficients between healthy and damaged data for each of the actuator-sensor path is used to calculate the field value for each pixel on the structure. Damage calibration curve is developed by progressively increasing the damage and obtaining a magnitude of the probability density function of the severity of the damage. Proposed approach has been validated using experimental data for multiple damage cases on plates, internal surfaces of pipes and impeller blade to successfully detect submillimeter scale holes and cracks, material adhesion as well as rate of pipeline erosion and corrosion.

Abstract We present a probabilistic rotor life prediction framework that combines the forging flaw crack nucleation process and the subsequent crack growth to failure. Experimental fatigue tests of specimens including forging flaws show that the life cycle of a forging flaw can be described by a nucleation phase followed by a fatigue crack growth phase. These results demonstrate that the nucleation phase is a significant fraction of the whole life cycle to failure. However, as there is no engineering method available that describes reliably the nucleation phase, this portion is oftentimes neglected in engineering life prediction frameworks, therefore resulting in a conservative life quantification. In order to improve probabilistic life quantification methods, we introduce a rigorous scheme that convolutes the local crack nucleation probabilities and the local crack growth failure probabilities in order to provide a local failure probability. Integration over the whole component then yields the total probability of failure for the engineering part under a specific load spectrum. A specific direct simulation Monte Carlo numerical implementation is demonstrated. It is applied to fatigue crack nucleation from large gas turbine rotor disk forging flaws followed by crack growth to component failure. For different regions of the analyzed rotor components, the results show the probabilistic interplay of the different temperature and stress dependences of the applied empirical nucleation models and the fatigue crack growth models. The presented probabilistic approach is generic and not restricted to the discussed fatigue nucleation and subsequent crack growth process in large rotor forgings. The framework can be applied to a variety of sequential failure processes including static and fatigue loading phenomena, as well as a multiplicity of failure modes and sequences relevant for engineering components.

Abstract In regions with high intermittent renewable energy share, thermal plants are forced to operate with greater flexibility beyond their original design intent. Decreasing energy prices and capacity factors will further force these plants to more transient operation with steeper load gradients. Older steam turbine (ST) protections systems on site are often not designed for such flexible operation and do not properly supervise the resulting impact on lifetime consumption. Therefore, precise lifetime management concepts are required to increase plant reliability and flexibility, and to mitigate risks for new implemented operation modes. Several lifetime assessment methods were developed to quantify the damage evolution and the residual lifetime for ST components. Usually, these methods require both input about representative or operated loading profiles as well as characteristic material curves. These characteristic curves are determined by a number of standardized material tests. Due to the material scatter and other different sources of uncertainties, each test is a realization of a stochastic process. Hence, the corresponding characteristic material curves inherit these uncertainties and do not represent an absolute limit. The analyses of different loading profiles even for the same plant and same start-up class reveal that the consideration of statistically evaluated specific start-up distributions and further transient events are of major importance. Probabilistic methods are able to quantify all of these uncertainties and compute the probability of failure for a given lifetime or vice versa. Within this paper, at first an extensive and systematic operational profile analysis is carried out and discussed, which acts as an input for a probabilistic lifetime assessment approach. For that, a developed probabilistic workflow is presented to quantify the uncertainties and for lifetime prediction using the generalized damage accumulation rule with focus on creep-fatigue loading. To quantify the characteristic material curves, existing experimental data of a 2%-Cr forged steel (23CrMoNiWV8-8) is used. A probabilistic representation of the Wilshire-Scharning equation characterizes the creep rupture behavior. The maximum-likelihood method is used for parameter estimation and to take still running long term creep experiments into account. The end of life in low cycle fatigue experiments is characterized by a macroscopic crack initiation and the Manson-Coffin-Basquin equation is utilized to represent the characteristic material curve. A temperature modified version of the Manson-Coffin-Basquin equation is used to represent the experimental data. The parameter estimation is done by using the linear regression analysis followed by a comprehensive regression diagnostic. Taking both, the material and the load scatter into account, a reliability analysis is carried out to compute the probability of crack initiation. Finally, different load cases are considered and evaluated against each other.

Abstract Operation and maintenance costs are a major driver for levelized cost of energy of wind power plants and can be reduced through optimized operation and maintenance practices accomplishable by various prognostics and health management (PHM) technologies. In recent years, the wind industry has become more open to adopting PHM solutions, especially those focusing on diagnostics. However, prognostics activities are, in general, still at the research and development stage. On the other hand, the industry has a request to estimate a component’s remaining useful life (RUL) when it has faulted, and this is a key output of prognostics. Systematically presenting PHM technologies to the wind industry by highlighting the RUL prediction need potentially helps speed up its acceptance and provides more benefits from PHM to the industry. In this paper, we introduce a PHM for wind framework. It highlights specifics unique to wind turbines and features integration of data and physics domain information and models. The output of the framework focuses on RUL prediction. To demonstrate its application, a data domain method for wind turbine gearbox fault diagnostics is presented. It uses supervisory control and data acquisition system time series data, normalizes gearbox temperature measurements with reference to environmental temperature and turbine power, and leverages big data analytics and machine-learning techniques to make the model scalable and the diagnostics process automatic. Another physics-domain modeling method for RUL prediction of wind turbine gearbox high-speed-stage bearings failed by axial cracks is also discussed. Bearing axial cracking has been shown to be the prevalent wind turbine gearbox failure mode experienced in the field and is different from rolling contact fatigue, which is targeted during the bearing design stage. The method uses probability of failure as a component reliability assessment and RUL prediction metric, which can be expanded to other drivetrain components or failure modes. The presented PHM for wind framework is generic and applicable to both land-based and offshore wind turbines.

This paper experimentally investigated the evolution of the tip clearance flow of a CRAC (Contra-Rotating Axial Compressor) test rig by means of high-response dynamic pressure measurements. The unsteady pressure field along both chordwise and circumferential directions in the tip clearance is recorded. The tip clearance vortex trajectory is captured using RMS (Root-Mean Square) method. Pressure spectrum analysis indicates that the unsteadiness of tip clearance vortex occurred when the flow coefficient approaches low enough even in the stable operating point. The unsteadiness of tip clearance vortex gets stronger as the flow coefficient drops until rotating stall occurs. According to this feature, the auto-correlation analysis and the cross-correlation analysis combined probability statistics method are used to work as pre-stall warning methods. In addition to, rotating instability which is caused by disturbances propagating along circumferential direction occurred at some flow condition.

In recent years, the demands of Maintenance, Repair and Overhaul (MRO) customers to provide resource-efficient after market services have grown increasingly. One way to meet these requirements is by making use of predictive maintenance methods. These are ideas that involve the derivation of workscoping guidance by assessing and processing previously unused or undocumented service data. In this context a novel approach on predictive maintenance is presented in form of a performance-based classification method for high pressure compressor (HPC) airfoils. The procedure features machine learning algorithms that establish a relation between the airfoil geometry and the associated aerodynamic behavior and is hereby able to divide individual operating characteristics into a finite number of distinct aero-classes. By this means the introduced method not only provides a fast and simple way to assess piece part performance through geometrical data, but also facilitates the consideration of stage matching (axial as well as circumferential) in a simplified manner. It thus serves as prerequisite for an improved customary HPC performance workscope as well as for an automated optimization process for compressor buildup with used or repaired material that would be applicable in an MRO environment. The methods of machine learning that are used in the present work enable the formation of distinct groups of similar aero-performance by unsupervised (step 1) and supervised learning (step 2). The application of the overall classification procedure is shown exemplary on an artificially generated dataset based on real characteristics of a front and a rear rotor of a 10-stage axial compressor that contains both geometry as well as aerodynamic information. In step 1 of the investigation only the aerodynamic quantities in terms of multivariate functional data are used in order to benchmark different clustering algorithms and generate a foundation for a geometry-based aero-classification. Corresponding classifiers are created in step 2 by means of both, the k Nearest Neighbor and the linear Support Vector Machine algorithms. The methods’ fidelities are brought to the test with the attempt to recover the aero-based similarity classes solely by using normalized and reduced geometry data. This results in high classification probabilities of up to 96 % which is proven by using stratified k-fold cross-validation.

Transpiration Cooling is an effective cooling technology to protect hot section components such as gas turbine airfoils, rocket heads and space craft. This external cooling method has much higher efficiency than film cooling with holes when consuming the same amount of coolant, due to the uniformity of coolant distribution. However, pore blockage, which frequently occur during the operation of transpiration cooled components, prevented its application in turbine components which require long term stability. Dust deposition was one the main reasons causing blockage of pores for transpiration cooling. A lot of effort was devoted into dust deposition and erosion while optimization for the components themselves were generally difficult as the blockage caused by dusts was unpredictable for traditional sintered porous media. Additive manufacturing, with capability to precisely construct structures in small scales, is a considerable tool to enhance the controllability of porous media, and furthermore, to find a good solution to minimize the blockage disadvantage. Present study selected a cooling configurations containing perforate straight holes with an additive manufacturable diameter of 0.4 mm. Computational Fluid Dynamics (CFD) methods were utilized to model the pore blockage and its effect on heat transfer. A scripting code in addition to the ANSYS CFX solver was utilized to simulate the random blockage conditions of the holes. Two hundred numerical cases with four different blockage probabilities were calculated and statistically evaluated to quantify the disadvantage of pore blockage on the cooling effectiveness. Results obtained from the numerical analysis indicated that the overall blockage ratio was a dominating parameter for the cooling effectiveness. Upstream regions of the cooled surface were more sensitive to local blockage compared to downstream regions. Randomness of the cooling effectiveness increased with the increase of blockage probability. Present study provided a quantitative understanding of the random blockage disadvantage on the specific transpiration cooling configuration, and could benefit further optimization effort to reduce the blockage disadvantage of transpiration cooling using additive manufacturing.

Performance changes due to manufacturing variations for a turbine blade are evaluated in the paper using the first order (FO) and second order (SO) sensitivities, which are calculated by a continuous adjoint method. In the case of newly manufactured blades, the geometric variation at each scanning point is assumed to meet a standard normal distribution. In the study, a modelling method taking into account the spatial correlations is employed to quantify the geometric uncertainty, where the contour of manufacturing tolerance is produced by imposing a series of shape functions on the blade. The basis modes of the geometric variations are extracted by principle component analysis. Firstly, the calculations of both FO and SO Sensitivities by solving the Euler adjoint equations are introduced. By regarding a finite number of primary basis modes of the geometric variations as the geometric parameters, the SO sensitivities can be obtained with significantly reduced computation cost. Sensitivity validations and performance evaluations based on sensitivities for each basis mode are then presented, illustrating the dramatic improvements on performance evaluations using the SO sensitivities. Finally, the statistics of performance changes for the turbine blade are evaluated by using the Monte Carlo simulations with respect to two different probability density functions for the input random variables. The results further demonstrate that the nonlinear dependence of the aerodynamic performance on the geometric variations can be captured by using the SO sensitivities.

This paper compares two back step film-cooling configurations under an uncertainty quantification framework. An important limit of such configurations is their reliability under geometrical variations, which is taken into account in this study. For the back step configurations, a straight and a curved step is used. Detached eddy simulations with k-ω turbulence model are performed using OpenFOAM ver. 4.0. The Reynolds number is based on the main stream velocity and film cooling hole diameter, d , and is Re = 15,300. The investigated step heights are 0.5 d and 0.75 d , and the blowing ratios, BR, are 0.5 and 1.0. The straight and the curved steps are found to enhance lateral spreading of coolant flow, resulting in higher film cooling effectiveness compared to the baseline case without the step at comparatively higher BR conditions. The curved step shows better performance than the straight one in particular from BR = 1.0 upwards with the step height of 0.5 d . At lower BR with lower H / d , and at higher BR with higher H / d , deterministic simulations are not able to identify the best performer. However when the performance of the two configurations is evaluated considering the stochastic variation of step height and the cooling condition, the benefit of the curved step becomes clear. In particular, the curved step shows better mean performance and has a higher probability to achieve a better performance than the other one. The uncertainty in the film cooling effectiveness caused by the uncertainty of the step height and the BR is investigated using Sparse Approximation of Moment-Based Arbitrary polynomial chaos (SAMBA).

A fan blade is a complicated object and obviously it is subjected to geometrical uncertainties from manufacture tolerances and other production deviations. In spite of all uncertainties a fan blade should provide stable aerodynamic efficiency and strength properties. That is why it is considered to solve a multidimensional and multidisciplinary optimization task (aerodynamics, strength and flutter sensitivity) in robust statement under geometrical uncertainties. In the proposed test case geometrical uncertainties from the fan blade manufacture tolerances and deviations are considered. The probability density function (pdf) was obtained as a result of statistical operation upon the results of blade coordinate measurements. Approximately 2500 fan blades were measured by means of CMM process to reconstruct the pdf for more than 40 geometrical uncertainties (there are blade thicknesses for different airfoil locations in several cross-sections). CFD and FEM calculations were carried out in NUMECA FINE/Turbo and ANSYS software, correspondingly. The surrogate model technique (the response surface and the Monte-Carlo method implemented to RSM results) was applied for the uncertainty quantification and the robust optimization process for the task under consideration. APPROX software was used for surrogate model construction. The IOSO technology was employed as one of the robust optimization tools. This technology is also based on a widespread application of the response surface technique. As a result, robust optimal solutions (the Pareto set) for all 4 considered criteria (aerodynamic efficiency, structural properties, stall margin and flutter sensitivity) were obtained. The probabilistic criteria were assessed based on the results obtained. The robust optimization results were compared with the deterministic optimization results.

Solid particle ingestion is one of the principal degradation mechanisms in the compressor and turbine sections of gas turbines. In particular, in industrial applications, the micro-particles not captured by the air filtration system can cause deposits on blades and, consequently, can result in a decrease in compressor performance. It is of great interest to the industry to determine which zones of the compressor blades are impacted by these small particles. However, this information often refers to single stage analysis. This paper presents three-dimensional numerical simulations of the micro-particle ingestion (0.15 μm – 1.50 μm) in a multistage (i.e. eight stage) subsonic axial compressor, carried out by means of a commercial CFD code. Particle trajectory simulations use a stochastic Lagrangian tracking method that solves the equations of motion separately from the continuous phase. The effects of humidity, or more generally, the effects of a third substance at the particle/surface interface (which is considered one of the major promoters of fouling) is then studied. The behavior of wet and oiled particles, in addition to the usual dry particles, is taken into consideration. In the dry case, the particle deposition is established only by using the sticking probability. This quantity links the kinematic characteristics of particle impact on the blade with the fouling phenomenon. In the other two cases, the effect of the presence of a third substance at the particle/surface interface is considered by means of an energy-based model. Moreover, the influence of the tangential impact velocity on particle deposition is analyzed. Introducing the effect of a third substance, such as humidity or oil, the phenomenon of fouling concerns the same areas of the multistage compressor. The most significant results are obtained by combining the effect of the third substance with the effect of the tangential component of the impact velocity of the particles. The deposition trends obtained with these conditions are comparable with those reported in literature, highlighting how the deposits are mainly concentrated in the early stages of a multistage compressor. Particular fluid dynamic phenomena, such as corner separations and clearance vortices, strongly influence the location of particle deposits.

Predictive lifing with probabilistic treatment of key variables represents a promising approach to realizing the digital gas turbine of the future. In this paper, we present a predictive model for creep life assessment of an uncooled turbine blade. The model development methodology draws on well-established machine learning principles to develop and validate a surrogate model for creep life from engine performance parameters. Verified creep life results, obtained from 3D non-linear thermo-mechanical finite element simulation for varying engine operating conditions are used as the basis for model development. The selection of model response surface order is studied over a range of models by evaluating normalized residual error on training and uncorrelated validation data sets. A model that is fully quadratic in the data set features is shown to have excellent predictive capability, yielding nominal creep life predictions to within ± 3% on the validation data set. This work then considers probabilistic techniques to evaluate the impact of uncertainty associated with each key factor on the predicted nominal creep life in order to achieve a mandated life target with a defined probability of failure.

Bladed Disks are subjected to different types of excitations, which cannot in any case be described in a deterministic manner. Fuzzy factors, such as slightly varying airflow or density fluctuation, can lead to an uncertain excitation in terms of amplitude and frequency, which has to be described by random variables. The computation of frictionally damped blades under random excitation becomes highly complex due to the presence of nonlinearities. Only a few publications are dedicated to this particular problem. Most of these deal with systems of only one or two degrees of freedom and use computational expensive methods, like finite element method (FEM) or finite differences method (FDM), to solve the determining differential equation. The stochastic stationary response of a mechanical system is characterized by the joint probability density function (JPDF), which is driven by the Fokker-Planck equation (FPE). Exact stationary solutions of the FPE only exist for a few classes of mechanical systems. This paper presents the application of a semi-analytical Galerkin -type method to a frictionally damped bladed disk under influence of Gaussian white noise (GWN) excitation in order to calculate its stationary response. One of the main difficulties is the selection of a proper initial approximate solution, which is applicable as a weighting function. Comparing the presented results with those from the FDM, Monte-Carlo Simulation (MCS) as well as analytical solutions proves the applicability of the methodology.