TASK Quarterly   Scientific Bulletin of the Centre of Informatics - Tricity Academic Supercomputer & networK   ISSN 1428-6394

1/2002 Computational Fluid Dynamics

Guest editor: dr Piotr Lampart, lampart@imp.gda.pl
Institute of Fluid Flow Machinery, Polish Academy of Sciences, Gdansk, Poland


From the History of Science and Technology in Ancient Gdansk
  • A.Januszajtis, The People of Gdansk: Fires and Electricity


  • J.D.Denton, The Effects of Lean and Sweep on Transonic Fan Performance a Computational Study

    The aerodynamics of transonic fans is discussed with emphasis on the use of three-dimensional design techniques, such as blade sweep and lean, to improve their performance. In order to study the interaction of these 3D features with the shock pattern a series of five different designs is produced and analysed by CFD. It is found that the 3D features have remarkably little effect on the shock pattern near the tip where the shock must remain perpendicular to the casing. Lower down the blade significant shock sweep, and hence reduced shock loss, can be induced by 3D design but this is usually at the expense of reduced stall margin and increased loss elsewhere along the blade span. Overall, very little change in efficiency is produced by blade sweep or lean. However, forwards lean of the rotor does produce a small increase in mass flow. Radial migration of the boundary fluid on the suction surface behind the shock is shown to play a large part in the aerodynamics near the blade tip.

  • M.Hoeger, U.Schmidt-Eisenlohr, S.Gomez, H.Sauer and R.Muller, Numerical Simulation of the Influence of a Fillet and a Bulb on the Secondary Flow in a Compressor Cascade

    Large fillet radii are typically found on highly loaded compressor rotors to ensure structural integrity. The objective of this paper is to investigate the impact of such real geometry effects on the flow at the hub section. Investigations were performed numerically for the idealised case of a plane compressor cascade with the 3D Navier-Stokes code TRACE_S. Realistic inlet boundary layer displacement thickness and typical loading levels close to stall are considered at low inlet Mach numbers Ma1 = 0.23. A large fillet with a relative radius of 16% chord length is considered as well as a 3D leading edge bulb-configuration designed at TU Dresden. The results are discussed in terms of iso-Mach surfaces, secondary flow patterns and spanwise incidence and turning. A complex 3D vortex system rises from the fillet radius, which improves the aerodynamic behaviour of the cascade at the end-wall section. With the bulb configuration the suction surface horse-shoe vortex leg was demonstrated to weaken the undesirable cross flow and by that to reduce the hazard of corner stall.

  • R.A.Van den Braembussche, Turbomachinery Component Design by Means of CFD

    A short overview of the main techniques for turbomachinery blade design based on CFD is followed by a more detailed description on an Optimisation- and Inverse Design method, developed at the von Karman Institute. The optimisation method uses an Artificial Neural Network to extract knowledge from a Database containing the results of previous designs and a Genetic Algorithm to define the optimum blade. The inverse design method makes use of the Euler or Navier-Stokes equations to predict how a given 3D blade shape should be modified to reach a prescribed pressure or Mach number distribution along the blade surface. Examples of transonic compressor and turbine blades, designed by both methods, illustrate the potential of these modern aero-design systems. Special attention is given to the problems related to existence and uniqueness and to those features that facilitate the practical use of these methods.

  • W.T.Tiow and M.Zangeneh, A Novel 3D Inverse Method for the Design of Turbomachinery Blades in Rotational Viscous Flow: Theory and Applications

    The development and application of a three-dimensional (3D) inverse methodology is presented for the design of turbomachinery blades. The design method is based on the specification of the blade loading distribution and the corresponding blade shape is systematically sought using directly the difference between the target and initial values. The design procedure comprises mainly of a CFD solver code and the blade-update algorithm to calculate the desired blade geometry as well as the corresponding 3D flow. The CFD code is a well-validated three-dimensional flow solver and has shock capturing ability to cope in both subsonic and high transonic-shocked, viscous flow. Fundamentally, it is a cell-vertex, finite volume, time-marching solver employing the multistage Runge-Kutta integrator in conjunction with accelerating techniques (local time stepping and grid sequencing). To account for viscosity, viscous forces are included in the solution using the log-law and mixing length models. The effects of rotating blades as well as tip clearance flow are also included in the flow prediction. The capabilities of the present method are demonstrated in the redesign of a transonic fan blade, the NASA Rotor 67. The redesign focuses on the shocked flow near the tip, where the effects of shock-boundary interaction and leakage flow are examined. The result shows conclusively that the shock-formation and its intensity in such a high-speed turbomachinery flow are well defined on the loading distributions. Simple guidelines to change the loading distribution can be followed using the proposed inverse methodology to improve the blade shape.

  • K.Denus and C.Osborne, Hydraulic Development of a Centrifugal Pump Impeller Using the Agile Turbomachinery Design System

    The impeller of an existing industrial pump (with both geometry and performance known) was analyzed and redesigned using an integrated, design/analysis, turbomachinery geometry modeling and flow simulation system. The purpose of the redesign was to achieve improved impeller performance (at the duty point). Fluid dynamics and geometry modeling parts of the design/analysis system were systematically applied: a) to analyse the existing impeller (impeller A), which was designed using conventional (routine in industry) hydraulic layout procedures, and b) to develop a new impeller (impeller B), using a coupled, multilevel 1D-Q3D-3D for the design and optimization. This paper discusses the features and advantages of the integrated design system, in which the coupled CAD/CFD approach is fully implemented. The analysis results are presented for impellers A and B, with the latter demonstrating a predicted increased efficiency and smaller size. Comparisons of the CFD results for both impellers reveal internal flow features that explain the improved impeller B performance levels.

  • P.Lampart and S.Yershov, 3D Shape Optimisation of Turbomachinery Blading

    The shape of HP gas and steam turbine stages, as well as of an LP exit stage of a steam turbine, is optimised numerically using a code Optimus and 3D RANS solver FLowER. The numerical method draws on direct constrained optimisation based on the method of deformed polyhedron. Values of the minimised objective function, that is stage losses with the exit energy are found from 3D viscous compressible computations. There are constraints imposed on the mass flow rate, exit swirl angle and reactions. Among the optimised parameters are stator and rotor blade numbers, stagger and twist angles, stator sweep and lean, both straight and compound. The optimisation gives new 3D designs with increased efficiencies.

  • A.Ruprecht, Unsteady Flow Simulation in Hydraulic Machinery

    In the field of hydraulic machinery Computational Fluid Dynamics (CFD) is routinely used today in research and development as well as in the daily design phase. Today in industry mostly steady state simulations are applied. In this paper, however, an overview of unsteady simulations is shown for different applications. The presented examples contain problems with self excited unsteadiness, vortex rope in the draft tube, as well as applications with externally forced unsteadiness by changing or moving geometries and rotor-stator interactions. For the shown applications the requirements, potential and limitations of unsteady flow analysis are assessed.

  • R.Eisinger and A.Ruprecht, Automatic Shape Optimisation of Hydro Turbine Components Based on CFD

    Since hydro turbines are designed individually according to the local situation, this requires a huge engineering effort. In order to reduce this effort, automatic optimisation tools are necessary. In this paper the shape optimisation of a turbine draft tube is shown. Different optimisation algorithms have been applied and will be discussed. From the used algorithms, the one based on approximated gradients seems to be the fastest.

  • K.Kozel and J.Fort, Modern Finite Volume Methods Solving Internal Flow Problems

    The lecture deals with explicit and implicit finite difference and mainly finite volume schemes that have been commonly used in last years for compressible and incompressible fluid flow problems. Mainly we mention schemes used for transonic flow computation (inviscid as well as viscous models) in internal and also external aerodynamics. We prefer modern schemes like RK multistage schemes, TVD and ENO schemes, implicit schemes or higher order schemes. Some results of numerical computation or numerical simulation are presented in the second part of the paper, including 2D and 3D transonic flow through a channel and a cascade (mainly of turbine type), 2D and 3D backward facing step flow (laminar and turbulent), 2D and 3D impinging jet flow (comparisons of some turbulent models). The last two cases have been computed only for incompressible viscous flows. Some tests for the efficiency of higher order scheme (4th-6th order) are presented for the backward facing step problem. In many examples we also present a comparison of numerical and experimental results.

  • J.E.Anker, J.F.Mayer and H.Stetter, A Preconditioned Solution Scheme for the Computation of Compressible Flow in Turbomachinery at Arbitrary Mach Numbers

    A preconditioned solution scheme for the computation of compressible flow in turbomachinery at arbitrary Mach numbers is presented. The preconditioning technique used is applied to a state-of-the-art explicit, time-marching Navier-Stokes code which originally was developed for compressible, high-speed turbomachinery applications. It combines the ideas of low Mach number preconditioning and artificial compressibility method into a unified approach where principally fluids with arbitrary equations of state can be simulated. As shown by the test cases presented, it allows the code to simulate flows efficiently and accurately independent of the Mach number. A description of the Navier-Stokes equations for rotating coordinate systems, along with the solution scheme and the details of the preconditioning method is given. Since turbomachinery computations are often performed on truncated domains, the solution scheme should be used in conjunction with non-reflecting boundary conditions. A change in the time-dependency of the equations due to preconditioning necessitates a modification of the boundary conditions. Thus, a derivation of the appropriate boundary conditions for the presented preconditioned scheme was performed and the resultant equations are given in this paper. The effectiveness of the new boundary conditions is demonstrated by comparing them with both boundary conditions that use the standard one-dimensional characteristic approach and the original boundary conditions for the non-preconditioned case.

  • V.Dolejsi, Numerical Simulation of Compressible Flow through Cascade of Profiles

    We deal with the numerical simulation of flow through a steam turbine. The system of the compressible Navier-Stokes equations accompanied by the state equation of perfect gas are numerically solved. We propose a new approach for the numerical solution of the Navier-Stokes equations, where the inviscid part is discretized by the finite volume method and the viscous part by the finite element method. To capture precisely the position of shock waves, the suitable mesh refinement method has to be applied. Therefore the anisotropic mesh adaptation technique equipped with some suitable modification for viscous compressible flow is used. The suitable application of the mentioned numerical adaptive method allows us to obtain sufficiently precise results without much requirement on CPU-time and computer memory.

  • L.Li, G.Li and Z.Feng, Numerical Simulation of Spontaneously Condensing Flows in a Plane Turbine Cascade

    The low efficiency of wet steam turbine is mainly attributed to wetness losses. To investigate the mechanisms which give rise to these losses, a fully Eulerian model has been developed for calculation of the wet steam flows with spontaneous condensation. In this model, the liquid phase is described with two conservation equations in Eulerian form and coupled with a solver of gas dynamics equations. With such a model, the existing code for simulation of single-phase flows can easily be changed to include wet steam two-phase flows in wet steam turbines. A numerical simulation of condensing flow in a plane turbine cascade is performed, and the numerical results are presented and compared with the experimental results.

  • T.Koronowicz, Rationality of Solving 3D Circulation Problems Exclusively with the Use of the Navier-Stokes Equation (Reynolds Equation)

    The paper presents problems connected with solving 3D circulation problems. The existence of vortex singularities in the flow behind a body is a characteristic feature of such problems. The vortices influence the velocity and pressure fields in the vicinity of a body. A proposition of rational merging the methods based on the N-S equation together with vortex methods (based on the vorticity equation) is discussed. Such a connection enables accurate and efficient determination of vortex singularities in the flow behind a body.