Two-Phase Flow Modeling and Measurements in Low-Pressure Turbines: Part 2 — Turbine Wetness Measurement and Comparison to CFD-Predictions

Author(s):  
M. Schatz ◽  
T. Eberle ◽  
M. Grübel ◽  
J. Starzmann ◽  
D. M. Vogt ◽  
...  

The correct computation of steam subcooling, subsequent formation of nuclei and finally droplet growth is the basic prerequisite for a quantitative assessment of the wetness losses incurred in steam turbines due to thermal and inertial relaxation. The same basically applies for the prediction of droplet deposition and the resulting threat of erosion. Despite the fact that there are many CFD-packages that can deal with real-gas effects in steam flows, the accurate and reliable prediction of subcooling, condensation and wet steam flow in steam turbines using CFD is still a demanding task. One reason for this is the lack of validation data for turbines that can be used to assess the physical models applied. Experimental data from nozzle and cascade tests can be found in the open literature; however, this data is only partly useful for validation purposes for a number of reasons. With regard to steam turbine test data, there are some publications, yet always without any information about the turbine stage geometries. This publication is part of a two-part paper; whereas part 1 focuses on the numerical validation of wet steam models by means of condensing nozzle and cascade flows, the focus in this part lies on the comparison of CFD results of the turbine flow to experimental data at various load conditions. In order to assess the validity and reliability of the experimental data, the method of measurement is presented in detail and discussed. The comparison of experimental and numerical results is used for a discussion about the challenges in both modeling and measuring steam turbine flows, presenting the current experience and knowledge at ITSM.

Author(s):  
M. Schatz ◽  
T. Eberle ◽  
M. Grübel ◽  
J. Starzmann ◽  
D. M. Vogt ◽  
...  

The correct computation of steam subcooling, subsequent formation of nuclei and finally droplet growth is the basic prerequisite for a quantitative assessment of the wetness losses incurred in steam turbines due to thermal and inertial relaxation. The same basically applies for the prediction of droplet deposition and the resulting threat of erosion. Despite the fact that there are many computational fluid dynamics (CFD)-packages that can deal with real-gas effects in steam flows, the accurate and reliable prediction of subcooling, condensation, and wet steam flow in steam turbines using CFD is still a demanding task. One reason for this is the lack of validation data for turbines that can be used to assess the physical models applied. Experimental data from nozzle and cascade tests can be found in the open literature; however, these measurement results are only partly useful for validation purposes for a number of reasons. With regard to steam turbine test data, there are some publications, yet always without any information about the turbine stage geometries. This publication is part of a two-part paper; whereas Part I focuses on the numerical validation of wet steam models by means of condensing nozzle and cascade flows, the focus in this part lies on the comparison of CFD results of the turbine flow to experimental data at various load conditions. In order to assess the validity and reliability of the experimental data, the method of measurement is presented in detail and discussed. The comparison of experimental and numerical results is used for a discussion about the challenges in both modeling and measuring steam turbine flows, presenting the current experience and knowledge at Institute of Thermal Turbomachinery and Machinery Laboratory (ITSM).


Author(s):  
Yasuhiro Sasao ◽  
Satoshi Miyake ◽  
Kenji Okazaki ◽  
Satoru Yamamoto ◽  
Hiroharu Ooyama

In this paper, we present an inclusive tracking algorithm for water droplets in a wet steam flow through a multi-stage steam turbine. This algorism is based on the Eulerian-Lagrangian coupled solver. The solver continuously computes water droplet growth, kinematic non-equilibrium between vapor and droplets, capture and kinetics of droplets on turbine blades, departure of large droplets from the trailing edge of blades, acceleration and atomization of large droplets, and recollisions between blades and droplets. Our Eulerian-Lagrangian coupled solver is used to predict wetness in unsteady three-dimensional (3D) wet steam flows through three-stage stator rotor cascade channels in a low pressure (LP) steam turbine model which is developed by Mitsubishi Heavy Industries (MHI). Droplet groups tracked by the discrete droplet model (DDM) are placed in the computational domain according to the predicted wetness. Interference from the gas phase on the droplets is considered, to track their kinetic and behavior, until they reach the outlet of the computational domain. The aim of this research is to investigate those multi-physics phenomena that trigger all forms of loss in steam turbines. In addition, this method will also be applied to multi-physics problems such as erosion in future work. This paper is presented as a first step in the research. Overviews of model of current coupling solver and several test calculations are presented.


2021 ◽  
Author(s):  
Nicola Maceli ◽  
Lorenzo Arcangeli ◽  
Andrea Arnone

Abstract Testing a sub-component or testing a scaled model are the approaches currently used to reduce the development cost of the new low-pressure (LP) section of a steam turbine. In any case, testing campaigns are run at a limited number of operating conditions. Therefore, some correlations are used to build a performance model of the LP module and expand the usage of a limited set of experimental data to cover the application range encountered in the steam turbine market. Another approach, which has become feasible during the last decade, is the usage of CFD calculations. These two approaches include a certain amount of uncertainty in the performance of the LP section, mainly related to the losses caused by the moisture content in the flow. In the present paper, the results of the analysis of a cutting-edge low-pressure section for small steam turbines are presented. The results are obtained by using a CFD commercial code with a set of user defined subroutines to model the effects of droplets nucleation and growth. Different operating conditions are considered, with different wetness at the exit and different pressure ratios, in order to clearly show the loss trend for different levels of exit moisture. The numerical results are compared with the experimental data, showing a significant improvement in the performance predictability for the considered case and demonstrating the benefit of using a CFD approach instead of using existing correlations.


Author(s):  
Tim Wittmann ◽  
Christoph Bode ◽  
Jens Friedrichs

Abstract This study investigates the applicability of an Euler-Lagrange approach for the calculation of nucleation and condensation of steam flows. Supersonic nozzles are used as generic validation cases, as their high expansion rates replicate the flow conditions in real turbines. Experimental and numerical validation data for these nozzles are provided by the International Wet Steam Modelling Project of Starzmann et al. (2018). In contrast to most participants of that project, an Euler-Lagrange approach is utilized for this study. Therefore, the classical nucleation theory with corrections and different droplet growth laws is incorporated into the Discrete Phase Model of ANSYS Fluent. Suggestions for an efficient implementation are presented. The Euler-Lagrange results show a good agreement with the experimental and numerical validation data. The sensitivities of the Euler-Lagrange approach to modelling parameters are analysed. Finally, an optimal parameter set for the calculation of nucleation and condensation is proposed.


Author(s):  
Yogini Patel ◽  
Giteshkumar Patel ◽  
Teemu Turunen-Saaresti

With the tremendous role played by steam turbines in power generation cycle, it is essential to understand the flow field of condensing steam flow in a steam turbine to design an energy efficient turbine because condensation at low pressure (LP) turbine introduces extra losses, and erosion in turbine blades. The turbulence has a leading role in condensing phenomena which involve a rapid change of mass, momentum and heat transfer. The paper presents the influence of turbulence modelling on non-equilibrium condensing steam flows in a LP steam turbine stage adopting CFD code. The simulations were conducted using the Eulerian-Eulerian approach, based on Reynolds-averaged Navier-Stokes equations coupled with a two equation turbulence model, which is included with nucleation and droplet growth model for the liquid phase. The SST k-ω model was modified, and the modifications were implemented in the CFD code. First, the performance of the modified model is validated with nozzles and turbine cascade cases. The effect of turbulence modelling on the wet-steam properties and the loss mechanism for the 3D stator-rotor stage is discussed. The presented results show that an accurate computational prediction of condensing steam flow requires the turbulence to be modelled accurately.


Author(s):  
Xiaofeng Zhu ◽  
Xin Yuan ◽  
Zhirong Lin ◽  
Naoki Shibukawa ◽  
Tomohiko Tsukuda ◽  
...  

The present paper proposes an Eulerian-Eulerian two-phase model for non-equilibrium condensing flow in steam turbines. This model is especially suitable for upwind finite volume scheme. An approximate Roe type flux using real water/vapor property is constructed to calculate the upwind wet-steam flux. This flux fully couples the wetness fraction with other conservative variables in the Jacobian Matrix whose eigen-vector and eigen-value are analitically derived. A novel treatment of real wet-steam property is developed by constructing a 3-DOFs TTSE table according to IAPWS97 formulas. The table is actually a cubic and uses the mixture’s density, the mixture’s internal energy and wetness as independent variables. Besides homogeneous condensation, heterogeneous condensing is also integrated into the model, which facilitates simulating the effect of salt impurities. The above methods are validated through two nozzle and one turbine cascade calculations and finally applied to a model LP steam turbine stage. Results show that the current model is very robust and is able to correctly capture the non-equilibrium condensation phenomena.


Author(s):  
Tadashi Tanuma ◽  
Yasuhiro Sasao ◽  
Satoru Yamamoto ◽  
Yoshiki Niizeki ◽  
Naoki Shibukawa ◽  
...  

The purpose of this paper is to present a numerical evaluation method for the aerodynamic design and development of high-efficiency exhaust diffusers in steam turbines, as well as to present the comparison between the numerical results and measured data in an actual real scale development steam turbine. This paper presents numerical investigation of three-dimensional wet steam flows in a down-flow-type exhaust diffuser that has non-uniform inlet flows from a typical last turbine stage. This stage has long transonic blades designed using recent aerodynamic and mechanical design technologies, including superimposed leakages and blade wakes from several upstream low pressure turbine stages. The present numerical flow analysis showed detail three-dimensional flow structures considering circumferential flow distributions caused by the down-flow exhaust hood geometry and the swirl velocity component from the last stage blades, including flow separations in the exhaust diffuser. The results were compared with experimental data measured in an actual development steam turbine. Consequently, the proposed aerodynamic evaluation method was proved to be sufficiently accurate for steam turbine exhaust diffuser aerodynamic designs.


Author(s):  
Christopher Fuhrer ◽  
Marius Grübel ◽  
Damian M. Vogt ◽  
Paul Petrie-Repar

Turbine blade flutter is a concern for the manufacturers of steam turbines. Typically, the length of last stage blades of large steam turbines is over one meter. These long blades are susceptible to flutter because of their low structural frequency and supersonic tip speeds with oblique shocks and their reflections. Although steam condensation has usually occurred by the last stage, ideal gas is mostly assumed when performing flutter analysis for steam turbines. The results of a flutter analysis of a 2D steam turbine test case which examine the influence of non-equilibrium wet steam are presented. The geometry and flow conditions of the test case are supposed to be similar to the flow near the tip in a steam turbine as this is where most of the unsteady aerodynamic work contributing to flutter is done. The unsteady flow simulations required for the flutter analysis are performed by ANSYS CFX. Three fluid models are examined: ideal gas, equilibrium wet steam (EQS) and non-equilibrium wet steam (NES), of which NES reflects the reality most. Previous studies have shown that a good agreement between ideal gas and EQS simulations can be achieved if the prescribed ratio of specific heats is equal to the equilibrium polytropic index of the wet steam flow through the turbine. In this paper the results of a flutter analysis are presented for the 2D test case at flow conditions with wet steam at the inlet. The investigated plunge mode normal to chord is similar to a bending mode around the turbine axis for a freestanding blade in 3D. The influence of the overall wetness fraction and the size of the water droplets at the inlet are examined. The results show an increase of aerodynamic damping for all investigated interblade phase angles with a reduction of droplet size. The influence of the wetness fraction is in comparison of less importance.


Author(s):  
Navid Sharifi ◽  
Majid Sharifi

Nucleation and condensation phenomena are of fundamental importance in many fields such as steam-turbine design and power generation technologies. Wet steam flows are typically considered as multiphase gas droplet mixtures in which both vapor and liquid droplets coexist. In such flows, spontaneous nucleation leads to the formation of liquid droplets from vapor. Our key goal is to determine the rate of nucleation and droplets growth correctly. This will enable us to predict the variations of thermodynamic properties along the nozzle axis. In this study, a CFD code is generated based on the assumption of non-isothermal homogenous nucleation rate. A “Used-Defined Function” (UDF) was written in a compatible format with FLUENT solver such that it implements all the required wet steam calculations through a finite-volume method. The predicted numerical results were well supported by experimental data from literature for a specific nozzle. The predicted distribution of pressure ratio along the main axis of the nozzle shows a reasonable agreement with experimental data. Moreover, the droplets sizes predicted were in good agreement with experimental data too. Besides, the variations of some important thermodynamic properties along the nozzles were determined as well. The predicted results were compared to available data from relevant literature. The outcome of current numerical procedure confirms the superiority of this module for wet steam considerations in supersonic flow. It can further be applied to wet flow analysis in so many applications such as steam turbine cascades.


Sign in / Sign up

Export Citation Format

Share Document