Table of contents

The operation of Francis turbines in part load condition causes high pressure fluctuations and dynamic loads in the turbine as well as high flow losses in the draft tube. Owing to the co-rotating velocity distribution at the runner blade trailing edge a low pressure zone arises in the hub region finally leading to a rotating vortex rope in the draft tube. A better understanding and a more accurate prediction of this phenomenon can help in the design process of a Francis turbine. The goal of this study is to reach a quantitatively better numerical prediction of the flow at part load and to evaluate the necessary numerical depth with respect to effort and benefit. As standard practice, simulation results are obtained for the steady state approach with SST turbulence modelling. Those results are contrasted with transient simulations applying a SST as well as a SAS (Scale Adaptive Simulation) turbulence model. The structure of the SAS model is such, that it is able to resolve the turbulent flow behaviour in more detail. The investigations contain a comparison of the flow losses in different turbine components.

A detailed flow evaluation is done in the cone and the diffuser of the draft tube. The different numerical approaches show a different representation of the vortex rope phenomenon indicating differences in pressure pulsations at different geometric positions in the entire turbine.

Finally, the turbulent flow structures in the draft tube are illustrated with several evaluation methods, such as turbulent eddy viscosity, velocity invariant and turbulent kinetic energy spectra.

An automatic Computational Fluid Dynamics (CFD) procedure that aims at predicting Draft Tube Pressure Pulsations (DTPP) at part load is presented. After a brief review of the physics involved, a description of the transient numerical setup is given. Next, the paper describes a post processing technique, namely the separation of pressure signals into synchronous, asynchronous and random pulsations. Combining the CFD calculation with the post-processing technique allows the quantification of the potential excitation of the mechanical system during the design phase. Consequently it provides the hydraulic designer with a tool to specifically target DTPP and thus helps in the development of more robust designs for part load operation of turbines.

In the framework of the BulbT project launched by the Consortium on Hydraulic Machines and the LAMH (Hydraulic Machine Laboratory of Laval University) in 2011, an intensive campaign to identify flow phenomena in the draft tube of a model bulb turbine has been done. A special focus was put on the draft tube component since it has a particular importance for recuperation in low head turbines. Particular operating points were chosen to analyse flow phenomena in this component. For each of these operating points, power, efficiency and pressure were measured following the IEC 60193 standard. Visualizations, unsteady wall pressure and efficiency measurements were performed in this component. The unsteady wall pressure was monitored at seven locations in the draft tube. The frequency content of each pressure signal was analyzed in order to characterize the flow phenomena across the efficiency hill chart. Visualizations were recorded with a high speed camera using tufts and cavitation bubbles as markers. The predominant detected phenomena were mapped and categorized in relation to the efficiency hill charts obtained for three runner blade openings. At partial load, the vortex rope was detected and characterized. An inflection in the partial load efficiency curves was found to be related to complex vortex rope instabilities. For overload conditions, the efficiency curves present a sharp drop after the best efficiency point, corresponding to an inflection on the power curves. This break off is more severe towards the highest blade openings. It is correlated to a flow separation at the wall of the draft tube. Also, due to the separation occurring in these conditions, a hysteresis effect was observed on the efficiency curves.

As part of the BulbT project, led by the Consortium on Hydraulic Machines and the LAMH (Hydraulic Machine Laboratory of Laval University), the efficiency and power break off in a bulb turbine has been investigated. Previous investigations correlated the break off to draft tube losses. Tuft visualizations confirmed the emergence of a flow separation zone at the wall of the diffuser. Opening the guide vanes tends to extend the recirculation zone. The flow separations were investigated with two-dimensional and two-component particle image velocimetry (PIV) measurements designed based on the information collected from tuft visualizations. Investigations were done for a high opening blade angle with a N11 of 170 rpm, at best efficiency point and at two points with a higher Q11. The second operating point is inside the efficiency curve break off and the last operating point corresponds to a lower efficiency and a larger recirculation region in the draft tube. The PIV measurements were made near the wall with two cameras in order to capture two measurement planes simultaneously. The instantaneous velocity fields were acquired at eight different planes. Two planes located near the bottom wall were parallel to the generatrix of the conical part of the diffuser, while two other bottom planes diverged more from the draft tube axis than the cone generatrix. The last four planes were located on the draft tube side and diverged more from the draft tube axis than the cone generatrix. By combining the results from the various planes, the separation zone is characterized using pseudo-streamlines of the mean velocity fields, maps of the Reynolds stresses and maps of the reverse-flow parameter. The analysis provides an estimation of the separation zone size, shape and unsteady character, and their evolution with the guide vanes opening.

Introduction of intermittent electricity production systems like wind power and solar systems to electricity market together with the consumption-based electricity production resulted in numerous start/stops, load variations and off-design operation of water turbines. The hydropower systems suffer from the varying loads exerted on the stationary and rotating parts of the turbines during load variations which they are not designed for. On the other hand, investigations on part load operation of single regulated turbines, i.e., Francis and propeller, proved the formation of rotating vortex rope (RVR) in the draft tube. The RVR induces oscillating flow both in plunging and rotating modes which results in oscillating force with two different frequencies on the runner blades, bearings and other rotating parts of the turbine. The purpose of this study is to investigate the effect of transient operations on the pressure fluctuations on the runner and mechanism of the RVR formation/mitigation. Draft tube and runner blades of the Porjus U9 model, a Kaplan turbine, were equipped with pressure sensors. The model was run in off-cam mode during different load variation conditions to check the runner performance under unsteady condition. The results showed that the transients between the best efficiency point and the high load happens in a smooth way while transitions to/from the part load, where rotating vortex rope (RVR) forms in the draft tube induces high level of fluctuations with two frequencies on the runner; plunging and rotating mode of the RVR.

For a prototype turbine operating under part load conditions, the turbine output power is fluctuating strongly. The test for the prototype turbine at site shows that the main reason is the resonance between the draft tube vortex frequency and the generator natural frequency. In order to reduce the fluctuation of power output, different measures are investigated with using CFD methods. To keep the turbine unchanged, four kinds of draft tubes are examined, including the original draft, the draft tube with extending runner cone, the draft tube with damping gates and the draft tube with flow deflectors. The results are analyzed and compared in order to examine the effects on pressure fluctuation and formation of vortex rope of draft tube. It is found that adding flow deflector is the most effective to change the frequency of the draft tube vortex rope and reduce the amplitude of pressure fluctuation.

The swirling flow in the draft tube of a Francis turbine can cause the flow instability and the cavitation surge and has a larger influence on hydraulic power operating system. In this paper, the cavitating flow with swirling flow in the diffuser was studied by the draft tube component experiment, the model Francis turbine experiment and the numerical simulation. In the component experiment, several types of fluctuations were observed, including the cavitation surge and the vortex rope behaviour by the swirling flow. While the cavitation surge and the vortex rope behaviour were suppressed by the aeration into the diffuser, the loss coefficient in the diffuser increased by the aeration. In the model turbine test the aeration decreased the efficiency of the model turbine by several percent. In the numerical simulation, the cavitating flow was studied using Scale-Adaptive Simulation (SAS) with particular emphasis on understanding the unsteady characteristics of the vortex rope structure. The generation and evolution of the vortex rope structures have been investigated throughout the diffuser using the iso-surface of vapor volume fraction. The pressure fluctuation in the diffuser by numerical simulation confirmed the cavitation surge observed in the experiment. Finally, this pressure fluctuation of the cavitation surge was examined and interpreted by CFD.

The decelerated swirling flow in the draft tube cone of hydraulic turbines (especially turbines with fixed blades) is responsible for self-induced instabilities which generates pressure pulsations that hinder the turbine operation. An experimental test rig was developed in order to investigate the flow instabilities. A new method was implemented to slow down the runner using a magneto rheological brake in order to be extended the flow regimes investigated. As a result, the experimental investigations are performed for 7 operating regimes in order to quantify the flow behaviour from part load operation to overload operation. The unsteady pressure measurements are carried out on 4 levels in the cone. The unsteady pressure measurements on the cone wall consist in quantifying of three aspects: i) the pressure recovery coefficient obtained based on mean pressure provides the energetic assessment on the draft tube cone; ii) the unsteady quantities (dominant amplitude and frequency) are determined revealing the dynamic behaviour; iii) the plunging and rotating components of the pressure pulsation. As a result, this new method helps us to investigate in detail the flow instability for different operating regimes and allows investigating various flow control solutions.

The flow unsteadiness generated in the draft tube cone of hydraulic turbines affects the turbine operation. Therefore, several swirling flow configurations are investigated using a swirling apparatus in order to explore the unsteady phenomena. The swirl apparatus has two parts: the swirl generator and the test section. The swirl generator includes two blade rows being designed such that the exit velocity profile resembles that of a turbine with fixed pitch. The test section includes a divergent part similar to the draft tube cone of a Francis turbine. A new control method based on a magneto rheological brake is used in order to produce several swirling flow configurations. As a result, the investigations are performed for six operating regimes in order to quantify the flow from part load operation, corresponding to runaway speed, to overload operation, corresponding to minimum speed, at constant guide vane opening. The part load operation corresponds to 0.7 times the best efficiency discharge, while the overload operation corresponds to 1.54 times the best efficiency discharge. LDV measurements are performed along three survey axes in the test section. The first survey axis is located just downstream the runner in order to check the velocity field at the swirl generator exit, while the next two survey axes are located at the inlet and at the outlet of the draft tube cone. Two velocity components are simultaneously measured on each survey axis. The measured unsteady velocity components are used to validate the results of unsteady numerical simulations, conducted using the OpenFOAM CFD code. The computational domain covers the entire swirling apparatus, including strouts, guide vanes, runner, and the conical diffuser. A dynamic mesh is used together with sliding GGI interfaces to include the effect of the rotating runner. The Reynolds averaged Navier-Stokes equations coupled with the RNG k-ε turbulence model are utilized to simulate the unsteady turbulent flow throughout the swirl generator.

In the BulbT project framework, a bulb turbine model was studied with a strongly diverging draft tube. At high discharge, flow separation occurs in the draft tube correlated to significant efficiency and power drops. In this context, a focus was put on the draft tube inlet flow conditions. Actually, a precise inlet flow velocity field is required for comparison and validation purposes with CFD simulation. This paper presents different laser Doppler velocimetry (LDV) measurements at the draft tube inlet and their analysis. The LDV was setup to measure the axial and circumferential velocity on a radius under the runner and a diameter under the hub. A method was developed to perform indirect measurement of the mean radial velocity component. Five operating conditions were studied to correlate the inlet flow to the separation in the draft tube. Mean velocities, fluctuations and frequencies allowed characterizing the flow. Using this experimental database, the flow structure was characterized. Phase averaged velocities based on the runner position allowed detecting the runner blade wakes. The velocity gradients induced by the blade tip vortices were captured. The guide vane wakes was also detected at the draft tube inlet. The recirculation in the hub wake was observed.

At both partial and full load of Francis turbines, the unsteady behavior of cavitating draft tube vortices occurs and leads to undesirable matters such as power house vibration, noise and power swing in some cases. This paper presents the investigation of the interaction between the flow pattern at runner outlet and the unsteady behavior of cavitating vortices in draft tube with experimental and numerical approaches. On the experimental research, the pressure pulsation in the draft tube is measured and the unsteady behavior of cavitating vortices is taken pictures with a high speed camera in the model test. On the numerical research, by Computational Fluid Dynamics (CFD) adopting a two-phase unsteady analysis, the analysis domain from the guide vane to the draft tube is carried out for investigating the interaction between the runner outlet flow and the vortex pattern. The pressure pulsation at the upper draft tube and the unsteady behavior of cavitating vortices obtained from CFD results are similar to those obtained in the model test. Detailed analysis of CFD results at overload indicates the repeat of expansion and contraction of cavitating vortices, which were shaped helical vortices with opposite direction of runner rotation, and the corresponding flow pattern in every time step of the pressure pulsations.

Runners with splitter blades were used widely for the high efficiency and stability. In this paper, the unsteady simulation of an ultra-high head turbine at the best efficiency point, 50% and 75% discharge points were established, to analyze the pressure pulsation in the vaneless space, rotating domain and the draft tube. First of all, runners with different length splitter blades and without splitter blades were compared to learn the efficiency and the pressure distribution on the blade surface. And then the amplitude of the pressure pulsation was analysed. The peak efficiency of the runner with splitter blades is remarkably higher than that of the corresponding impeller without splitter blades. And the efficiency of the turbine is the highest when the length ratio of the splitter blades is 0.75 times the main blades. The pressure pulsation characteristics were also influenced, because the amplitudes of the pulsation induced by the RSI phenomenon were changed as a result of more blades. At last, the best design plan of the length of the splitter blades (length ratio=0.825) was obtained, which improved the pressure pulsation characteristics without significant prejudice to the efficiency.

Francis turbines have been running more and more frequently in part load conditions, in order to satisfy the new market requirements for more dynamic and flexible energy generation, ancillary services and grid regulation. The turbines should be able to be operated for longer durations with flows below the optimum point, going from part load to deep part load and even speed-no-load. These operating conditions are characterised by important unsteady flow phenomena taking place at the draft tube cone and in the runner channels, in the respective cases of part load and deep part load. The current expectations are that new Francis turbines present appropriate hydraulic stability and moderate pressure pulsations at overload, part load, deep part load and speed-no-load with high efficiency levels at normal operating range. This study presents series of investigations performed by Voith Hydro with the objective to improve the hydraulic stability of Francis turbines at overload, part load and deep part load, reduce pressure pulsations and enlarge the know-how about the transient fluid flow through the turbine at these challenging conditions. Model test measurements showed that distinct runner designs were able to influence the pressure pulsation level in the machine. Extensive experimental investigations focused on the runner deflector geometry, on runner features and how they could reduce the pressure oscillation level. The impact of design variants and machine configurations on the vortex rope at the draft tube cone at overload and part load and on the runner channel vortex at deep part load were experimentally observed and evaluated based on the measured pressure pulsation amplitudes. Numerical investigations were employed for improving the understanding of such dynamic fluid flow effects. As example for the design and experimental investigations, model test observations and pressure pulsation curves for Francis machines in mid specific speed range, around n_qopt = 50 min⁻¹ are reported, analysed and commented here. The analysis of experimental results allowed the identification of designs and configurations, which can improve the machine stability.

A measurement campaign using unsteady wall pressure sensors on a bulb turbine draft tube was performed over the power and efficiency break off range of a N₁₁ curve. This study is part of the BulbT project, undertaken by the Consortium on hydraulic machines and the LAMH (Hydraulic Machine Laboratory of Laval University). The chosen operating points include the best efficiency point for a high runner blade angle and a high N₁₁. Three other points, with the same N₁₁, have been selected in the break off zone of the efficiency curve. Flow conditions have been set using the guide vanes while the runner blade angle remained constant. The pressure sensors were developed from small piezoresistive chips with high frequency response. The calibration gave an instrumental error lower than 0.3% of the measurement range. The unsteady wall pressure was measured simultaneously at 13 locations inside the first part of the draft tube, which is conical, and at 16 locations in the circular to rectangular transition part just downstream. It was also measured at 11 locations along a streamwise line path at the bottom left part of the draft tube, where flow separation occurs, covering the whole streamwise extent of the draft tube. For seven radial-azimuthal planes, four sensors were distributed azimuthally.

As confirmed by tuft visualizations, the break off phenomenon is correlated to the presence of flow separation inside the diffuser at the wall. The break off is linked to the appearance of a large recirculation in the draft tube. The efficiency drop increases with the size of the separated region. Analysis of the draft tube pressure coefficients confirms that the break off is related to diffuser losses. The streamwise evolution of the mean pressure coefficient is analyzed for the different operating conditions. An azimuthal dissymmetry of the mean pressure produced by the separation is detected. The pressure signals have been analyzed and used to track the separation zone depending on the operating conditions. Spectral analysis of these signals reveals a low frequency unsteadiness generated by the flow separation.

Although the technology of pump-turbines is generally well known the operation is still affected by flow phenomena that are quite complex and not fully understood. One of these phenomena is the S-shape instability which occurs in turbine mode at low load operation, close to runaway conditions. The instability results in an S-shape of the turbine characteristics and complicates the synchronization of the machine. Numerical investigations performed in the past indicated that the occurrence of turbine instabilities is connected with the appearance of rotor-stator interactions, and backflow regions in the vane less space between guide vane and impeller. This paper presents the results and conclusions of experimental investigations of pump-turbine instabilities carried out to find a practical explanation for the flow phenomena responsible for the appearance of the S-shaped characteristics. In the scope of a joint research project with Andritz Hydro, the Institute for Hydraulic Fluidmachinery at Graz University of Technology optimized an existing 4-quadrant test rig for an experimental investigation at off design conditions featuring the possibility for adjusting stable operation of instabilities. All the experimental investigations were based on IEC60193-standard using a pump turbine model provided by Andritz Hydro AG. In addition to the standard measurements of flow rate, head and efficiency the interaction between model and its hydraulic environment were analysed by dynamic pressure sensors. Additional pressure sensors integrated in the guide vane apparatus were used to investigate pressure distributions in the model. Particle Image Velocimetry (PIV) allowed the measurement of the velocity field in the vane less space between impeller and guide vanes and in the environment of two single guide vanes. The experimental investigations were focused on operation points in the S-shape region of the characteristics. For each operation point 190 double images for 20 rotor-stator positions were taken which allowed an analysis of a complete blade channel. The combination of PIV and pressure measurements in the model enabled a structured experimental analysis of the flow phenomena at low load off- design operation and allowed an improved understanding of the physical background of the occurrence of the instability in turbine mode.

Hydropower is already the largest single renewable electricity source today but its further development will face new deployment constraints such as large-scale projects in emerging economies and the growth of intermittent renewable energy technologies. The potential role of hydropower as a grid stabilizer leads to operating hydro power plants in "off-design" zones. As a result, new methods of analyzing associated unsteady phenomena are needed to improve the design of hydraulic turbines. The key idea of the development is to compute a spatial description of a phenomenon by using a combination from several sensor signals. The spatial harmonic decomposition (SHD) extends the concept of so-called synchronous and asynchronous pulsations by projecting sensor signals on a linearly independent set of a modal scheme. This mathematical approach is very generic as it can be applied on any linear distribution of a scalar quantity defined on a closed curve. After a mathematical description of SHD, this paper will discuss the impact of instrumentation and provide tools to understand SHD signals. Then, as an example of a practical application, SHD is applied on a model test measurement in order to capture and describe dynamic pressure fields. Particularly, the spatial description of the phenomena provides new tools to separate the part of pressure fluctuations that contribute to output power instability or mechanical stresses. The study of the machine stability in partial load operating range in turbine mode or the comparison between the gap pressure field and radial thrust behavior during turbine brake operation are both relevant illustrations of SHD contribution.

The storage pumps are equipped with various types of inlet casings. The flow nonuniformity is generated by the suction elbows being ingested by the impeller leading to unsteady phenomena and worse cavitational behaviour. A symmetrical suction elbow model corresponding to the double flux storage pump was manufactured and installed on the test rig in order to assess the flow field at the pump inlet. The experimental investigations are performed for 9 discharge values from 0.5 to 1.3 of nominal discharge. LDV measurements are performed on the annular section of the pump inlet in order to quantify the flow non-uniformity generated by the symmetrical suction elbow. Both axial and circumferential velocity components are simultaneously measured on the half plane (180°) of the annular inlet section along to 19 survey axis with 62 points on each. The flow field on the next half plane is determined tacking into account the symmetry. As a result, the flow map on the pump inlet annular section is reconstructed revealing a significant variation of the circumferential velocity component. The absolute flow angle is computed showing a significant variation of ±38°.

The unsteady non-cavitation and cavitation turbulent flows in a centrifugal pump at partial load condition are numerically investigated by CFX 13.0. The numerical framework employs the combination of RNG k-ε turbulence model and transport equation cavitation model, in which the effects of compressibility of fluid on cavitation region and pressure fluctuation on saturation pressure are both taken into consideration. The good agreement between the numerical and experimental values validates that the numerical framework can accurately predict the turbulent flows in the centrifugal pump. The complex flow characteristics in impeller at non-cavitation and cavitation conditions are revealed. For the noncavitation flow, the dominant frequencies of pressure fluctuation of monitoring points in impeller are all the Impeller Rotation Frequency 24.17Hz. The maximum value of pressure fluctuation on the blade pressure side appears at the 0.8 chord length from the blade leading edge due to a clockwise rotating vortex, which incepts, develops and disappears when the corresponding blade passes through the volute tongue. The dominant frequencies of pressure fluctuation of monitoring points in volute are the Blade Pass Frequency 145 Hz or twice of it. The maximum value of pressure fluctuation in the volute appears near the tongue region, where the flow fields are uneven with strong second flow in the cross section. For the cavitation flow, as the cavitation develops at the blade leading edge, the turbulent flows in the impeller are greatly influenced by the bubble shedding and collapse. The maximum values of pressure fluctuation in impeller increase with the development of cavitation, and reach the largest magnification of about 2.0 in comparison to the non-cavitation flow when the pressure at the pump inlet is very low. The complicated phenomenon of unsteady turbulent flow in a centrifugal pump indicates that the vortex has great influence on the flow pattern.

A detailed measurement campaign is performed in which the effect is determined of modifications to the pump compartment on the velocity profile in the suction pipe of a vertically submersible pump. These measurements are performed in a scale model of a simple pumping station of which the dimensions are equal to the ANSI/HI 9.8-2012 guidelines (HIS standard). A set of generally accepted modifications like splitters, fillings, curtain walls and baffles has been tested. The investigation showed that the acceptability of a velocity profile is not significantly influenced by these modifications (if well-designed). Some modifications have a local effect on the velocity profile. However, the locations with the largest spatial deviation, which determine the acceptability of the velocity profile, are not affected. Therefore, the statement in the HIS standard that small modifications to the pump compartment improve the velocity profile seems to be not generically valid.

Rotating stall may occur at part load flow of a pump-turbine in pump mode. Unstable flow structures developing under stall condition can lead to a sudden drop of efficiency, high dynamic load and even cavitation. CFD simulations on a pump-turbine model in pump mode were carried out to reveal the onset and developed mechanisms of these unstable flow phenomena at part load. The simulation results of energy-discharge and efficiency characteristics are in good agreement with those obtained by experiments. The more deviate from design conditions with decreasing flow rate, the more flow separations within the vanes. Under specific conditions, four stationary separation zones begin to progress on the circumference, rotating at a fraction of the impeller rotation rate. Rotating stalls lead to the flow in the vane diffuser channels alternating between outward jet flow and blockage. Strong jets impact the spiral casing wall causing high pressure pulsations. Severe separations of the stall cells disturb the flow inducing periodical large amplitude pressure fluctuations, of which the intensity at different span wise of the guide vanes is different. The enforced rotating nonuniform pressure distributions on the circumference lead to dynamic uniform forces on the impeller and guide vanes. The results show that the CFD simulations are capable to gain the complicated flow structure information for analysing the unstable characteristics of the pump mode at part load.

This paper presents various numerical setups for modelling Francis turbine startups involving moving meshes and variable runner speed in order to help define best practices. During the accelerating phase of the startup, the flow is self-similar between channels, thus making single sector configuration appropriate. Adding the draft tube improves the results by allowing pressure recovery midway during in the startup. At the speed no-load regime, a rotating stall phenomenon occurs and can only be capted with the full runner included in the simulation. Comparison with experimental data, such as runner speed and strain gauge measurements, generally shows good agreement.

The turbine startup is a transient event generating high amplitude stress cycles in runner blades which might reduce significantly the life expectancy of the runner. Our goal is to use the increased range of possibilities offered by newer governing system to optimize the wicket gates control pattern in order to reduce the amplitude of the strain transient in the runner blade during startup. In this paper, we present our success in defining an optimal wicket gates control pattern for the startup of a newly commissioned propeller runner.

The main objective of the work presented in this paper is to investigate numerically the flow behavior inside a Francis hydro-turbine during the transient event of load rejection. First, a theoretical description of the flow during the event is presented in order to predict the global flow characteristics to be anticipated since no velocity profiles are available for this transient event. The issue of choosing the proper boundary conditions to obtain the absolute pressure and the correct flow characteristics within the runner when using a typical truncated geometry is then discussed. Finally, by using a hypothesis of "quasi-stationarity" and a validated methodology, global flow characteristics within the turbine are highlighted near the no-load operating condition and the unsteady vortical motions within the runner are assessed.

The causes of resonance of a certain model pump-turbine unit during startup process were investigated in this article. A three-dimensional full flow path analysis model which contains spiral case, stay vanes, guide vanes, runner, gaps outside the runner crown and band, and draft tube was constructed. The transient hydraulic excitation force of full flow path was analyzed under five conditions near the resonance region. Based on one-way fluid- structure interaction (FSI) analysis model, the dynamic stress characteristics of the pump-turbine runner was investigated. The results of pressure pulsation, vibration mode and dynamic stress obtained from simulation were consistent with the test results. The study indicated that the hydraulic excitation frequency (Z_g*f_n) Hz due to rotor-stator interference corresponding to the natural frequency of 2ND+4ND runner mode is the main cause of resonance. The relationship among pressure pulsation, vibration mode and dynamic stress was discussed in this paper. The results revealed the underlying causes of the resonance phenomenon.

Typically, transients such as load rejection generate only a few high vibration cycles in Francis runners. However, in the cases presented in this study, a sustained vibration around a natural frequency was observed on three (3) homologous Francis runners of different sizes during such events. The first two (2) runners were equipped with strain gauges on the blades and displacement sensors positioned circumferentially in the bottom ring and head cover around the runner labyrinth seals. The third runner was monitored only with displacement sensors on non-rotating components. The data from the first two (2) runners provided a better understanding of the parameters influencing the appearance of the high amplitude vibrations and allowed the implementation of a test plan to circumvent the phenomenon during commissioning of the third runner. Based on the measured data, the distributor's closing parameters were optimized to eliminate the vibration observed during load rejection on most of the operating range and reduce it significantly at full load.

Recent researches indicate that the useful life of a turbine can be affected by transient events. This study aims to define and validate strategies for the simulation of the flow within a propeller turbine model in runaway condition. Using unsteady pressure measurements on two runner blades for validation, different strategies are compared and their results analysed in order to quantify their precision. This paper will focus on justifying the choice of the simulations strategies and on the analysis of preliminary results.

Runaway speed is an important performance factor for the safe operation of hydropower systems. In turbine design, the manufacturers must conduct several model tests to calculate the accurate value of runaway speed for the complete range of operating conditions, which are expensive and time-consuming. To study runaway conditions, the application of numerical tools such as unsteady CFD simulations can help to better understand the complex flow physics during transient processes. However, unsteady simulations require significant computational effort to compute accurate values of runaway speed due to difficulties related to unsteady turbulent flow modelling and instabilities. The present study presents a robust methodology based on steady-state RANS flow simulations capable of predicting the runaway speed of a Francis turbine with an adequate level of accuracy and in a reasonable simulation time. The simulations are implemented using a commercial flow solver and an iterative algorithm that relies on a smooth relation between turbine torque and speed coefficient. The impact of friction has been considered when estimating turbine torque, in order to improve the accuracy. The results of this study show good agreement with experiments.

During the start-up process, the strong instabilities happened at no load operation in a low head axial flow fixed-blade turbine, with strong pressure pulsation and vibration. The rated speed can not reach until guide vane opening to some extent, and stable operation could not be maintained under the rated speed at some head, which had a negative impact on the grid-connected operation of the unit. In order to find the reason of this phenomenon, the unsteady flow field of the whole flow passage at no load conditions was carried out to analyze the detailed fluid field characteristics including the pressure pulsation and force imposed on the runner under three typical heads. The main hydraulic cause of no load conditions instability was described. It is recommended that the power station should try to reduce the no-load running time and go into the high load operation as soon as possible when connected to grid at the rated head. Following the recommendations, the plant operation practice proved the unstable degree of the unit was reduced greatly during start up and connect to the power grid.

Modern reversible pump-turbines can start in turbine operation very quickly, i.e. within few minutes. Unfortunately no clear design rules for runners with a stable start-up are available, so that certain machines can present unstable characteristics which lead to oscillations in the hydraulic system during synchronization. The so-called S-shape, i.e. the unstable characteristic in turbine brake operation, is defined by the change of sign of the slope of the head curve. In order to assess and understand this kind of instabilities with CFD, fast and reliable methods are needed. Using a 360 degrees model including the complete machine from spiral casing to draft tube the capabilities of a newly developed in-house tool are presented. An ad-hoc simulation is performed from no-load conditions into the S-shape in transient mode and using moving-mesh capabilities, thus being able to capture the opening process of the wicket gates, for example like during start-up. Beside the presentation of the computational methodology, various phenomena encounterd are analyzed and discussed, comparing them with measured and previously computed data, in order to show the capabilities of the developed procedure. Insight in detected phenomena is also given for global data like frequencies of vortical structures and local flow patterns.

The runaway process in a model pumped-storage system was simulated for analyzing the dynamic characteristics of a pump-turbine. The simulation was adopted by coupling 1D (One Dimensional) pipeline MOC (Method of Characteristics) equations with a 3D (Three Dimensional) pump-turbine CFD (Computational Fluid Dynamics) model, in which the water hammer wave in the 3D zone was defined by giving a pressure dependent density. We found from the results that the dynamic performances of the pump-turbine do not coincide with the static operating points, especially in the S-shaped characteristics region, where the dynamic trajectories follow ring-shaped curves. Specifically, the transient operating points with the same Q₁₁ and M₁₁ in different moving directions of the dynamic trajectories give different n₁₁. The main reason of this phenomenon is that the transient flow patterns inside the pump-turbine are influenced by the ones in the previous time step, which leads to different flow patterns between the points with the same Q₁₁ and M₁₁ in different moving directions of the dynamic trajectories.

In this paper, a new nonlinear k-ε turbulence model based on RNG k-ε turbulence model and Wilcox's k-ω turbulence model was proposed to simulate the unsteady flow and to predict the pressure fluctuation through a model pump turbine for engineering application. Calculations on a curved rectangular duct proved that the nonlinear k-ε turbulence model is applicable for high pressure gradient flows and large curvature flows. The numerically predicted relative pressure amplitude (peak to peak) in time domain to the pump turbine head at no load condition is very close to the experimental data. It is indicated that the prediction of the pressure fluctuation is valid by the present nonlinear k-ε method. The high pressure fluctuation in this area is the main issue for pump turbine design, especially at high head condition.

The upper partial load unsteady phenomena are often observed at model tests for Francis turbine with high and middle specific speed. It is appears approximately between 7085% of optima point discharge for constant unit speed value and has accompanied by additional phenomenon with much higher frequency than draft tube vortex precession frequency and also runner rotational frequency. There are some discussions about nature of this phenomena and transposition of unsteady model test results to the prototype. In this paper are presented the results of above mentioned phenomena model investigations and some results of investigation at prototype turbine. Based on the results of model tests the following extensive data have been obtained: pressure fluctuation in the draft tube cone and spiral case, axial force fluctuations, it is demonstrated the significant influence of cavitation on upper partial load unsteady phenomena. The result of measurements of bearing vibrations and pressure pulsations are presented for prototype turbine at corresponded or very close operation points to model. In accordance with obtained data it is demonstrated that at upper partial load operation the unsteady phenomenon is observed as for the model also for the prototype turbine. On the base of model investigation has been demonstrated the influence of air admission and special design solutions to diminish unsteady phenomena at upper partial load range. All investigations were based on the physical experiment. Thus, based on model and prototype experimental investigations it is obtained additional information about upper partial load unsteady phenomenon and confirmed the transposition of model results to prototype turbine.

At high load operation points, Francis turbines generally produce large cavitation volumes of central vortex character in the draft tube. In order to gain a deeper understanding of the flow behaviour at high load conditions a combined 1D-3D transient two-phase numerical investigation at prototype size was carried out and these results were compared with measured site data. A one-dimensional model to capture hydroacoustic effects along a pipeline will be presented. The corresponding PDEs were solved using an implicit finite difference scheme on a staggered grid. In contrast to previous studies this model is coupled to the commercial software ANSYS CFX through an interface which exchanges pressure and discharge data within every time step until convergence. Results of the one-dimensional approach as well as the coupled solution were validated with commercial one-dimensional software (SIMSEN) and a full threedimensional calculation for hydroacoustic test cases. Unlike former investigations the described 1D-3D approach is used to compare site data with a numerical analysis at prototype size focused on the amplitude and frequency of the pressure pulsation at overload condition. The combined model is able to capture the occurring phase change in the draft tube as well as the propagating pressure oscillation through the hydraulic system without solving for the whole penstock in a 3D manner, thus saving time and computational resources.

The required operating range for hydraulic machines is continually extended in an effort to integrate renewable energy sources with unsteady power outputs into the existing electrical grid. The off-design operation however brings forth unfavorable flow patterns in the machine, causing dynamic problems involving cavitation, which may represent a limiting factor to the energy production. In Francis turbines it is observed that the self-excited oscillation of a vortex rope in the draft tube cone prevents the delivery of maximum power when required. This phenomenon is referred to as full load pressure surge and has been the object of extensive research during the past decades. Several contributions deepened its understanding through measurement and simulation of the local flow properties and the global stability parameters. The draft tube pressure level and the runner outlet swirl are identified as key variables in the modelling of the vortex rope dynamics. Recently, a cyclic appearance of blade cavitation has been observed at overload conditions in a multiphase numerical simulation coupling the runner and the draft tube. From the analysis of the simulation it becomes obvious that the cyclic appearance of blade cavitation has a direct effect on the runner outlet swirl, thus introducing an additional interaction mechanism that is not accounted for in formerly published models. For the presented work, the results of this numerical study are confirmed experimentally on a reduced scale model of a Francis turbine. Several wall pressure measurements in the draft tube cone are performed, together with high speed visualizations of the vortex rope and the blade cavitation. The flow swirl is calculated based on Laser Doppler Velocimetry measurements. A possible mechanism explaining the coupling between the self-excited pressure and vortex rope oscillation and the cyclic appearance of the blade cavitation is proposed. Furthermore, the streamwise propagation speed of the flow swirl in the draft tube is calculated. The results offer important insights in the physics of high load pressure surge and contribute to the further development of numerical draft tube flow and stability models.

Francis turbines operating at part load condition experience the development of a cavitating helical vortex rope in the draft tube cone at the runner outlet. The precession movement of this vortex rope induces local convective pressure fluctuations and a synchronous pressure pulsation acting as a forced excitation for the hydraulic system, propagating in the entire system. In the draft tube, synchronous pressure fluctuations with a frequency different to the precession frequency may also be observed in presence of cavitation. In the case of a matching between the precession frequency and the synchronous surge frequency, hydro-acoustic resonance occurs in the draft tube inducing high pressure fluctuations throughout the entire hydraulic system, causing torque and power pulsations. The risk of such resonances limits the possible extension of the Francis turbine operating range. A more precise knowledge of the phenomenon occurring at such resonance conditions and prediction capabilities of the induced pressure pulsations needs therefore to be developed.

This paper proposes a detailed study of the occurrence of hydro-acoustic resonance for one particular part load operating point featuring a well-developed precessing vortex rope and corresponding to 64% of the BEP. It focuses particularly on the evolution of the local interaction between the pressure fluctuations at the precession frequency and the synchronous surge mode passing through the resonance condition. For this purpose, an experimental investigation is performed on a reduced scale model of a Francis turbine, including pressure fluctuation measurements in the draft tube and in the upstream piping system. Changing the pressure level in the draft tube, resonance occurrences are highlighted for different Froude numbers. The evolution of the hydro-acoustic response of the system suggests that a lock-in effect between the excitation frequency and the natural frequency may occur at low Froude number, inducing a hydro-acoustic resonance in a random range of cavitation numbers.

The two-phase 1D-3D model of full load surge, presented previously by the authors, is applied to simulation of self-excited oscillations in prototype power station. The model consists in 1D hydro-acoustic equations for the penstock domain, and 3D two-phase unsteady RANS equations for turbine domain, including the cavitating flow in the draft tube. Both systems of equations are linked together by pressure and discharge at the penstock-turbine interface and solved simultaneously. To demonstrate the work of the model over-load pressure pulsations are simulated for high head prototype power plant. The influence of numerical parameters is investigated. It was shown that the choice of cavitation model has a strong effect on the amplitude of pulsations. The structure of the vapor cavity is investigated. The computed frequency is compared to the results of 1D model and with experimental data. Amplitudes and frequencies of the pulsations agree well with experiment, showing the potential of the method to predict full-load surge.

Hydraulic machines operating in a wider range are subjected to cavitation developments inducing undesirable pressure pulsations which could lead to potential instability of the power plant. The occurrence of pulsating cavitation volumes in the runner and the draft tube is considered as a mass source of the system and is depending on the cavitation compliance. This dynamic parameter represents the cavitation volume variation with the respect to a variation of pressure and defines implicitly the local wave speed in the draft tube. This parameter is also decisive for an accurate prediction of system eigen frequencies. Therefore, the local wave speed in the draft tube is intrinsically linked to the eigen frequencies of the hydraulic system. Thus, if the natural frequency of a hydraulic system can be determined experimentally, it also becomes possible to estimate a local wave speed in the draft tube with a numerical model.

In the present study, the reduced scale model of a Francis turbine (v=0.29) was investigated at off-design conditions. In order to measure the first eigenmode of the hydraulic test rig, an additional discharge was injected at the inlet of the hydraulic turbine at a variable frequency and amplitude to excite the system. Thus, with different pressure sensors installed on the test rig, the first eigenmode was determined. Then, a hydro-acoustic test rig model was developed with the In-house EPFL SIMSEN software and the local wave speed in the draft tube was adjusted to obtain the same first eigen frequency as that measured experimentally. Finally, this method was applied for different Thoma and Froude numbers at part load conditions.

A reversible pump turbine is a machine that can operate in three modes of operation i.e. in pumping mode. in turbine mode and in phase compensating mode (idle speed). Reversible pump turbines have an increasing importance for regulation purposes for obtaining power balance in electric power systems. Especially in grids dominated by thermal energy. reversible pump turbines improve the overall power regulating ability. Increased use of renewables (wind-. wave- and tidal power plants) will utterly demand better regulation ability of the traditional water power systems. enhancing the use of reversible pump turbines. A reversible pump turbine is known for having incredible steep speed - flow characteristics. As the speed increases the flow decreases more than that of a Francis turbines with the same specific speed. The steep characteristics might cause severe stability problems in turbine mode of operation. Stability in idle speed is a necessity for phasing in the generator to the electric grid. In the design process of a power plant. system dynamic simulations must be performed in order to check the system stability. The turbine characteristics will have to be modelled with certain accuracy even before one knows the exact turbine design and have measured characteristics. A representation of the RPT characteristics for system dynamic simulation purposes is suggested and compared with measured characteristics. The model shows good agreement with RPT characteristics measured in The Waterpower Laboratory. Because of the S-shaped characteristics. there is a stability issue involved when measuring these characteristics. Without special measures. it is impossible to achieve stable conditions in certain operational points. The paper discusses the mechanism when using a throttle to achieve system stability. even if the turbine characteristics imply instability.

Water hammer phenomena are important issues for high head hydro power plants. Especially, if several reversible pump-turbines are connected to the same waterways there may be strong interactions between the hydraulic machines. The prediction and coverage of all relevant load cases is challenging and difficult using classical simulation models. On the basis of a recent pump-storage project, dynamic measurements motivate an improved modeling approach making use of the Thoma number dependency of the actual turbine behaviour. The proposed approach is validated for several transient scenarios and turns out to increase correlation between measurement and simulation results significantly. By applying a fully automated simulation procedure broad operating ranges can be covered which provides a consistent insight into critical load case scenarios. This finally allows the optimization of the closing strategy and hence the overall power plant performance.

The hump region of head-discharge characteristic curve has great influence on the operation stability of pump-turbine. Therefore, it is significant to conduct deep research on the flow characteristic in the hump district of pump mode. Detailed study of the internal flow in hump district has been made combined with model experiments in this paper. Research discussed the adaptability of different turbulent model toward calculation in pump mode conditions and carried on three-dimensional numerical simulation of the whole pump-turbine flow passage. Then the head, efficiency and power characteristic curves of different pump modes were analysed. The numerical simulation results are in good agreement with experimental data. Comparing flow field both in hump district and of optimal operating point, it is found that the formation of hump area is concerned with complex flow in the runner and guide vanes domains, such as secondary flow, backflow and vortex patterns. The study aims to explore the cause for hump district generating of pump-turbine and its internal flow mechanism.

The measurements and video observation of unsteady flow in the draft tube cone of the pump-turbine model were conducted in the Laboratory of Water Turbines, property of OJSC "Power machines" - "LMZ". The prototype head was about 250 m. The experiments were performed for the turbine mode of operation.

Measurements were taken for the unit speed value n11 corresponding to rated head in the generating mode of operation, for a wide range of guide vanes openings at loads ranging from partial to maximum value. The researches of the velocity field in function of the Thoma number were carried out in some operating conditions.

The mean values and RMS deviations of the velocity components were the results of laser measurements. The curves of the intensity of the vortex versus the guide vane opening and the Thoma number were plotted. The energy velocity spectra were presented for the points at which the most pronounced frequency precession of the helical axial vortex was observed. Video recording and laser Doppler anemometry were made in the operating conditions of the developed cavitation. Based on the results of video observations and energy spectra obtained via LDA, vortex frequencies were determined i.e. the frequencies of the vortex precession under the runner in the draft tube cone.

The dynamic response of submerged and confined disk-like structures is of interest in the flied of hydraulic machinery, especially in hydraulic turbine runners. This response is difficult to be estimated with accuracy due to the strong influence of the boundary conditions. Small radial gaps as well as short axial distances to rigid surfaces greatly modify the dynamic response because the fact of the added mass and damping effects. Moreover, the effect of the shaft coupling is also important for certain mode-shapes of the structure. In the present study, the influence of the added mass effect and boundary conditions on the dynamic behavior of a submerged disk attached to a shaft is evaluated through experimental tests and structural- acoustic coupling numerical simulations. For the experimentation, a test rig has been developed. It consists of a confined disk attached to a shaft inside a cylindrical container full of water. The disk can be fixed at different axial positions along the shaft. Piezoelectric patches are used to excite the disk and the response is measured with submersible accelerometers. For each configuration tested, the natural frequencies of the disk and the shaft are studied. Numerical results have been compared with experimental results.

To study the dynamic behavior of turbine runners (natural frequencies and mode shapes) not only the added mass effect of still water has to be considered. Also the effect of rotation may not be neglected in the dynamic response. In the present study, the dynamic behavior of a rotating disk submerged in water is studied. For this purpose an experimental test rig has been developed. It consists of a rotating disk submerged in water that can be excited and its response can be measured from the rotating system by a slip ring system. For the excitation an impact device installed on the casing has been used. The response is measured with miniature accelerometers screwed on the disk. The influence of rotation on the dynamic response has been determined experimentally.

Fluid-structure interaction (FSI) has a major impact on the dynamic response of the structural components of hydroelectric turbines. On mid-head to high-head Francis runners, the rotor-stator interaction (RSI) phenomenon always has to be considered carefully during the design phase to avoid operational issues later on. The RSI dynamic response amplitudes are driven by three main factors: (1) pressure forcing amplitudes, (2) excitation frequencies in relation to natural frequencies and (3) damping. The prediction of the two first factors has been largely documented in the literature. However, the prediction of fluid damping has received less attention in spite of being critical when the runner is close to resonance. Experimental damping measurements in flowing water on hydrofoils were presented previously. Those results showed that the hydro-dynamic damping increased linearly with the flow. This paper presents development and validation of a mathematical model, based on momentum exchange, to predict damping due to fluid structure interaction in flowing water. The model is implemented as an analytical procedure for simple structures, such as cantilever beams, but is also implemented in more general ways using three different approaches for more complex structures such as runner blades: a finite element procedure, a CFD modal work based approach and a CFD 1DOF approach. The mathematical model and all three implementation approaches are shown to agree well with experimental results.

Fluid - Structure Interaction (FSI) phenomena is becoming a relevant study field for the design or revamping of hydropower plants. The generalized trend of increasing flow rates and reducing rotor blades/stay vanes thickness in order to improve the efficiency of the machine together with a major push from plant owners/operators for production flexibility (partial load operation is more common nowadays) make the FSI between the vortex shedding phenomenon and the vanes/blades of the machine an area of interest. From a design point of view, the machine structure has to resist all the hydrodynamic forces generated and maintain tension stresses under the fatigue limit to ensure a machine lifetime of several decades. To accomplish that goal, designers have to assure there is no presence of strong coupling phenomena (lock-in) between the vortex shedding frequency and the eigenfrequencies of the structure.

As the vortex street is directly related to the state of the boundary layer along the hydrofoil, in this paper the effect of the boundary layer on the vortex shedding in a Donaldson-type hydrofoil is studied using Computational Fluid Dynamics (CFD). The development of the boundary layer along the Donaldson trailing edge hydrofoil chord is presented under lock-off conditions. The results are validated against previously obtained experimental results. Since the Donaldson trailing edge is non-symmetric, the boundary layer velocity profiles are reported for the suction and pressure side of the hydrofoil. In addition, the effect of the Donaldson trailing edge on laminar-to-turbulent transition on both sides of the hydrofoil is studied.

The changes in the electricity market have led to changed requirements for the operation of pump turbines. Utilities need to change fast and frequently between pumping and generating modes and increasingly want to operate at off-design conditions for extended periods. Operation of the units in instable areas of the machine characteristic is not acceptable and may lead to self-excited vibration of the hydraulic system. In turbine operation of pump turbines unstable behaviour can occur at low load off-design operation close to runaway conditions (S-shape of the turbine characteristic). This type of instability may impede the synchronization of the machine in turbine mode and thus increase start-up and switch over times. A pronounced S-shaped instability can also lead to significant drop of discharge in the event of load rejection. Low pressure on the suction side and in the tail-race tunnel could cause dangerous separation of the water column. Understanding the flow features that lead to the instable behaviour of pump turbines is a prerequisite to the design of machines that can fulfil the growing requirements relating to operational flexibility. Flow simulation in these instability zones is demanding due to the complex and highly unsteady flow patterns. Only unsteady simulation methods are able to reproduce the governing physical effects in these operating regions. ANDRITZ HYDRO has been investigating the stability behaviour of pump turbines in turbine operation in cooperation with several universities using simulation and measurements. In order to validate the results of flow simulation of unstable operating points, the Graz University of Technology (Austria) performed detailed experimental investigations. Within the scope of a long term research project, the operating characteristics of several pump turbine runners have been measured and flow patterns in the pump turbine at speed no load and runaway have been examined by 2D Laser particle image velocimetry (PIV). For several wicket gate positions, the flow fields in the vane-less space at runner inlet observed in the experiment are compared with the results of unsteady CFD flow simulations. Physical phenomena are visualized and insight to flow phenomena is given. Analyses using both results of simulation and measurement allow deriving a consistent explanation of the fluid mechanical mechanisms leading to the S-shaped instability of pump turbines.

Pump-turbines were always running at partial condition with the power grid changing. Flow separations and stall phenomena were obvious in the pump-turbine. Most of the RANS turbulence models solved the shear stress by linear difference scheme and they were isotropous, so they couldn't capture all kinds of vortexes in the pump-turbine well. At present, Partially-Averaged Navier-Stokes (PANS) has been found better than LES in simulating flow regions especially those with poor near-wall resolution. In this paper, a new nonlinear PANS turbulence model was proposed, which was modified from RNG k-ε turbulence model and the shear stresses were solved by Ehrhard's nonlinear methods. The nonlinear PANS model was used to study the instability of "S" region of a model pump-turbine with misaligned guide vanes (MGV). The opening of pre-opened guide vanes had great influence on the "S" characteristics. Pressure fluctuations in the vaneless space for different opening of pre-opened guide vanes were analyzed. It is found that the "S" characteristics and instability can be improved when the relative pre-opening of MGV is 50%.

Flexibility and energy storage are one of the main challenges of the energy industry at the present time. Pumped Storage Power Plants (PSP), using reversible pump-turbines, are among the most cost-efficient solutions to answer these needs. To provide a rapid adjustment to the electricity grid, pump-turbines are subject of quick switching between pumping and generating modes and to extended operation under off-design conditions. In particular, at part load, instabilities in pump characteristics can occur. It can lead to unsteadiness and even to a shift of the operating point with significant modification of discharge and drop of efficiency. This unstable area is often exposed to the cavitation phenomenon, which can lead to vibrations, loss of performance and sometimes erosion. The paper focuses on the numerical analysis of the pumping mode regime, especially on the part load off-design instabilities, observed as a saddle shaped pump-turbine head curve and the presence and development of the cavitation in the part load area. The investigations were made on the reduce-scaled model of a high head pump-turbine design. Numerical calculations were performed using commercial code with implemented barotropic cavitation model. Some of the numerical results were compared to the experimental data. Flow analysis was stressed on the cavitation influence on the flow behavior and the performance of the machine. The analysis was made for various flow rates and a wide range of NPSH values. The importance of specific parts of the numerical domain for obtained results was investigated and evaluated.

Similarity study of prototype and model pumped turbines are performed by numerical simulation and the partial discharge case is analysed in detail. It is found out that in the RSI (rotor-stator interaction) region where the flow is convectively accelerated with minor flow separation, a high level of similarity in flow patterns and pressure fluctuation appear with relative pressure fluctuation amplitude of model turbine slightly higher than that of prototype turbine. As for the condition in the runner where the flow is convectively accelerated with severe separation, similarity fades substantially due to different topology of flow separation and vortex formation brought by distinctive Reynolds numbers of the two turbines. In the draft tube where the flow is diffusively decelerated, similarity becomes debilitated owing to different vortex rope formation impacted by Reynolds number. It is noted that the pressure fluctuation amplitude and characteristic frequency of model turbine are larger than those of prototype turbine. The differences in pressure fluctuation characteristics are discussed theoretically through dimensionless Navier-Stokes equation. The above conclusions are all made based on simulation without regard to the penstock response and resonance.

In a pumped storage power station, the units produce vibration and noise at times when the guide vanes rotate into the slight opening region during the turbine operating mode. According to this phenomenon, the simulation of transient flow in the units during the motion of the guide vane is carried out to investigate the variation of flow state in the process of startup and shutdown in turbine mode. The changing rate of hydraulic torque on a single guide vane is introduced to quantitatively represent the varying acuteness of the flow in the guide vanes and the possibility of the noise induced by the instable flow. The correlation between the frequency of noise and water head is summarized. The research indicates that the repeating reversal of fluid after load rejection is the hydraulic phenomenon which is the cause of the distributor vibration and noises within the slight opening, which is in accordance with the data recorded during the operation of the station. The effect of guide vanes closing law on the flow state in guide vanes and hydraulic torque on a single guide vane is analyzed.

When starting up a reversible Francis pump-turbine in pump mode, the machine may operate at zero flow at a given gate opening. Besides reversal flow and prerotation in the draft tube cone, the onset of a fully separated flow in the vaned diffuser is observable at zero- discharge condition. In this paper, the occurrence of prerotation and reversal flow in the conical draft tube and the flow in one stay vane channel of a pump-turbine are examined experimentally and compared to numerical simulations. In order to assess the strongly three-dimensional flow in the stay vane channel, measurements with a 2D laser doppler velocimeter (LDV) were performed at various positions. The inlet flow in the draft tube cone, which becomes significantly at zero discharge in pump mode, is investigated by velocity measurements at two different positions. Pressure fluctuations in the draft tube cone induced by complex flow patterns are also recorded and analyzed. It is found that the swirl number at zero discharge does not significant differ from the values obtained at very low load pumping. Experimental investigations combined with CFD have shown that in the stay vane channel flow velocity components different from zero occur even at no discharge. Streamline plots show the fully separated flow structure.

More efforts are put on hydro-power to balance voltage and frequency within seconds for primary control in modern smart grids. This requires hydraulic turbines to run at off-design conditions. especially at low load or speed-no load. Besides. the tendency of increasing power output and decreasing weight of the turbine runners has also led to the high level vibration problem of the runners. especially high head Francis runners. Therefore. it is important to carry out the static and dynamic stress analyses of prototype high head Francis runners. This paper investigates the static and dynamic stresses on the prototype high head Francis runner based on site measurements and numerical simulations. The site measurements are performed with pressure transducers and strain gauges. Based on the measured results. computational fluid dynamics (CFD) simulations for the flow channel from stay vane to draft tube cone are performed. Static pressure distributions and dynamic pressure pulsations caused by rotor-stator interaction (RSI) are obtained under various operating conditions. With the CFD results. static and dynamic stresses on the runner at different operating points are calculated by means of the finite element method (FEM). The agreement between simulation and measurement is analysed with linear regression method. which indicates that the numerical result agrees well with that of measurement. Furthermore. the maximum static and dynamic stresses on the runner blade are obtained at various operating points. The relations of the maximum stresses and the power output are discussed in detail. The influences of the boundary conditions on the structural behaviour of the runner are also discussed.

The Francis turbine runners of the Grimsel 2 pump storage power plant showed repeatedly cracks during the last decade. It is assumed that these cracks were caused by flow induced forces acting on blades and eventual resonant runner vibrations lead to high stresses in the blade root areas. The eigenfrequencies of the runner were simulated in water using acoustic elements and compared to experimental data. Unsteady blades pressure distribution determined by a transient CFD simulation of the turbine were coupled to a FEM simulation. The FEM simulation enabled analyzing the stresses in the runner and the eigenmodes of the runner vibrations. For a part-load operating point, transient CFD simulations of the entire turbine, including the spiral case, the runner and the draft tube were carried out. The most significant loads on the turbine runner resulted from the centrifugal forces and the fluid forces. Such forces effect temporally invariant runner blades loads, in contrast rotor stator interaction or draft tube instabilities induce pressure fluctuations which cause the temporally variable forces. The blades pressure distribution resulting from the flow simulation was coupled by unidirectional-harmonic FEM simulation. The dominant transient blade pressure distribution of the CFD simulation were Fourier transformed, and the static and harmonic portion assigned to the blade surfaces in the FEM model. The evaluation of the FEM simulation showed that the simulated part load operating point do not cause critical stress peaks in the crack zones. The pressure amplitudes and frequencies are very small and interact only locally with the runner blades. As the frequencies are far below the modal frequencies of the turbine runner, resonant vibrations obviously are not excited.

Reliable fatigue life assessment of Francis runners combines two parts: At first, the load universe describing how the plant will be operated. And secondly, for all essential operating conditions, component stresses due to static and dynamic loading have to be predicted and considered in the design process by the manufacturer. Therefore, dynamic loading conditions and the resulting impact on the fatigue life of hydroelectric components are an integral part of research activities. Especially off-design conditions and transient operations have been addressed in the last years. Based on strain gauge measurements in prototype runners, model test experiences, and advanced numerical simulations, the understanding of dynamic loads has been highly improved. From correlations of measurement and simulation, standard procedures have been developed to enhance the fatigue life. The present paper summarizes findings of recent investigations enabling Francis runners which combine high efficiency and a robust mechanical design.

Some of the potentially most damaging continuous operating conditions for hydraulic turbines are the no-load (NL) conditions. At NL conditions the flow passes through the turbine without power generation, but with non-negligible flow rate, and therefore all the potential energy in the flow has to be dissipated. This takes place through a mechanism where the runner channels are partially pumping, thus generating large scale unsteady vortex structures which, by their nature, break down into smaller and smaller vortices until energy dissipation occurs at the smallest scales. This type of flow, dominated by its turbulent character, is inherently difficult to simulate by means of numerical methods since turbulence model and numerical dissipation have a major influence. The resulting dynamic loads on the runner are largely of stochastic nature, exciting a broad band of frequencies and thus, almost always interact with at least one deformation mode. The presented investigations are aimed at predicting the effect of the unsteady NL pressure loads on the fatigue life of a Francis turbine runner. A combination of computational fluid dynamics (CFD) and finite element analysis (FEA) methods has been employed. The results from transient CFD simulations are presented. Comparison of the results with prototype strain gauge measurements at no load conditions shows that the stochastic nature and the approximate range of the dynamic stresses can be predicted.

The frequency of the Nordic power grid has become more volatile during recent years. This gives rise to two effects in synchronous machines. Firstly, as grid frequency changes the rotational speed of synchronous machines must change likewise. Secondly, the Nordic grid uses speed droop operation extensively as the primary governing; hence the power produced is a function of the grid frequency. These two coupled effects will lead to the runner and axle on synchronous machines having to cope with a varying level of torque. Even if the unit is supposedly operating at steady state via a fixed set point for the production, the influence of the varying grid frequency is that the torque in not steady at all. Recent years' new high head Francis runners in Norway have shown a tendency towards experiencing fatigue to a greater extent than what seem to be the case for new runners decades ago. Leading to this paper, measurements have been made of the rotational speed; generator power; main servo motor position and grid frequency at a Francis turbine unit. Based on these measurements simulations that include the hydraulic domain have then been performed. From these simulation results a property is constructed which is intended as a qualitative measure of the material stresses induced in the rotating masses of the unit, and is representative of the dynamical loads on the material of the rotating masses. The work is a part of a longer term goal, namely identifying the stress oscillations in a Francis turbine runner operating at speed of rotation oscillating because of grid frequency variations.