Compressor Blade Optimization Using a Continuous Adjoint Formulation

2006 ◽  
Author(s):  
Dimitiros I. Papadimitriou ◽  
Kyriakos C. Giannakoglou
Keyword(s):  
Stokes Equations ◽  
Compressor Blade ◽  
Navier Stokes ◽  
Computational Domain ◽  
Navier Stokes Equations ◽  
Adjoint Formulation ◽  
Design Variables ◽  
Gradient Computation ◽  
Geometrical Constraints ◽  
Continuous Adjoint

In this paper, a constrained optimization algorithm is formulated and utilized to improve the aerodynamic performance of a 3D peripheral compressor blade cascade. The cascade efficiency is measured in terms of entropy generation along the developed flowfield, which defines the field objective functional to be minimized. Its gradient with respect to the design variables, which are the coordinates of the Non-Uniform Rational B-Spline (NURBS) control points defining the blade, is computed through a continuous adjoint formulation of the Navier-Stokes equations based on the aforementioned functional. The steepest descent algorithm is used to locate the optimal set of design variables, i.e. the optimal blade shape. In addition to the well-known advantages of the adjoint method, the current formulation has even less CPU cost for the gradient computation as it leads to gradient expression which is free of field variations in geometrical quantities (such as derivatives of interior grid node coordinates with respect to the design variables); the computation of the latter would be costly since it requires remeshing anew the computational domain for each bifurcated design variable. The geometrical constraints, which depend solely on the blade parameterization, are handled by a quadratic penalty method by introducing additional Lagrange multipliers.

Shape Optimization of Forward-Curved-Blade Centrifugal Fan with Navier-Stokes Analysis

2004 ◽  
Vol 126 (5) ◽  
pp. 735-742 ◽  
Author(s):  
Kwang-Yong Kim ◽  
Seoung-Jin Seo
Keyword(s):  
Response Surface ◽  
Stokes Equations ◽  
Computing Time ◽  
Relative Size ◽  
Three Dimensional ◽  
Navier Stokes ◽  
Centrifugal Fan ◽  
Navier Stokes Equations ◽  
Design Variables ◽  
Curved Blade

In this paper, the response surface method using a three-dimensional Navier-Stokes analysis to optimize the shape of a forward-curved-blade centrifugal fan is described. For the numerical analysis, Reynolds-averaged Navier-Stokes equations with the standard k-ε turbulence model are discretized with finite volume approximations. The SIMPLEC algorithm is used as a velocity–pressure correction procedure. In order to reduce the huge computing time due to a large number of blades in forward-curved-blade centrifugal fan, the flow inside of the fan is regarded as steady flow by introducing the impeller force models. Four design variables, i.e., location of cutoff, radius of cutoff, expansion angle of scroll, and width of impeller, were selected to optimize the shapes of scroll and blades. Data points for response evaluations were selected by D-optimal design, and a linear programming method was used for the optimization on the response surface. As a main result of the optimization, the efficiency was successfully improved. Effects of the relative size of the inactive zone at the exit of impeller and momentum fluxes of the flow in scroll on efficiency were further discussed. It was found that the optimization process provides a reliable design of this kind of fan with reasonable computing time.

Numerical Investigation of Blade Flutter at or Near Stall in Axial Flow Turbomachines

2001 ◽  
Author(s):  
Wolfgang Höhn
Keyword(s):  
Viscous Flow ◽  
Stokes Equations ◽  
Separated Flow ◽  
Parallel Computer ◽  
Navier Stokes ◽  
Computational Domain ◽  
Compressor Cascade ◽  
Governing Equations ◽  
Navier Stokes Equations ◽  
Blade Flutter

During the design of the compressor and turbine stages of today’s aeroengines, aerodynamically induced vibrations become increasingly important since higher blade load and better efficiency are desired. In this paper the development of a method based on the unsteady, compressible Navier-Stokes equations in two dimensions is described in order to study the physics of flutter for unsteady viscous flow around cascaded vibrating blades at stall. The governing equations are solved by a finite difference technique in boundary fitted coordinates. The numerical scheme uses the Advection Upstream Splitting Method to discretize the convective terms and central differences discretizing the viscous terms of the fully non-linear Navier-Stokes equations on a moving H-type mesh. The unsteady governing equations are explicitly and implicitly marched in time in a time-accurate way using a four stage Runge-Kutta scheme on a parallel computer or an implicit scheme of the Beam-Warming type on a single processor. Turbulence is modelled using the Baldwin-Lomax turbulence model. The blade flutter phenomenon is simulated by imposing a harmonic motion on the blade, which consists of harmonic body translation in two directions and a rotation, allowing an interblade phase angle between neighboring blades. Non-reflecting boundary conditions are used for the unsteady analysis at inlet and outlet of the computational domain. The computations are performed on multiple blade passages in order to account for nonlinear effects. A subsonic massively stalled unsteady flow case in a compressor cascade is studied. The results, compared with experiments and the predictions of other researchers, show reasonable agreement for inviscid and viscous flow cases for the investigated flow situations with respect to the Steady and unsteady pressure distribution on the blade in separated flow areas as well as the aeroelastic damping. The results show the applicability of the scheme for stalled flow around cascaded blades. As expected the viscous and inviscid computations show different results in regions where viscous effects are important, i.e. in separated flow areas. In particular, different predictions for inviscid and viscous flow for the aerodynamic damping for the investigated flow cases are found.

scholarly journals PARALLEL COMPUTATIONS IN STUDIES OF DEPENDENCE OF GAS DYNAMIC PARAMETERS OF UPWARD SWIRLING FLOW OF GAS ON BLOWING VELOCITY

2016 ◽  
pp. 92-97
Author(s):  
R. E. Volkov ◽  
A. G. Obukhov
Keyword(s):  
Stokes Equations ◽  
Swirling Flow ◽  
Initial Conditions ◽  
Exact Analytical Solution ◽  
Three Dimensional ◽  
Navier Stokes ◽  
Heat Conducting ◽  
Computational Domain ◽  
Navier Stokes Equations ◽  
Gas Dynamic

The rectangular parallelepiped explicit difference schemes for the numerical solution of the complete built system of Navier-Stokes equations. These solutions describe the three-dimensional flow of a compressible viscous heat-conducting gas in a rising swirling flows, provided the forces of gravity and Coriolis. This assumes constancy of the coefficient of viscosity and thermal conductivity. The initial conditions are the features that are the exact analytical solution of the complete Navier-Stokes equations. Propose specific boundary conditions under which the upward flow of gas is modeled by blowing through the square hole in the upper surface of the computational domain. A variant of parallelization algorithm for calculating gas dynamic and energy characteristics. The results of calculations of gasdynamic parameters dependency on the speed of the vertical blowing by the time the flow of a steady state flow.

Front Tracking Approach and Application to Multi-Fluid Flow

2006 ◽  
Author(s):  
Manasa Ranjan Behera ◽  
K. Murali
Keyword(s):  
Stokes Equations ◽  
Navier Stokes ◽  
Computational Domain ◽  
Navier Stokes Equations ◽  
Flow Problems ◽  
Fixed Grid ◽  
Moving Interface ◽  
Fluid Properties ◽  
Volume Method ◽  
Transfer Of Information

Multiphase flows simulations using a robust interface-tracking method, are presented. The method is based on writing one set of governing equations for the whole computational domain and treating the different phases as single fluid domain with variable material properties. Interfacial terms are accounted for by adding the appropriate sources as δ functions at the boundary separating the phases. The unsteady Navier-Stokes equations are solved by finite volume method on a fixed, structured grid and the interface, or front, is tracked explicitly by a lower dimensional grid. Interfacial source terms are computed on the front and transferred to the fixed grid. Advection of fluid properties such as density and viscosity is done by following the motion of the front. The method has been implemented for interfacial flow problems, depicting the interface and topology change capturing capability. The representation of the moving interface and its dynamic restructuring, as well as the transfer of information between the moving front and the fixed grid, is discussed. Extensions of the method to density stratified flows, and interfacial movements are then presented.

Numerical Simulations of Nonlinear Post-Critical Vibrations of an Airfoil After Flutter Instability

2011 ◽  
Author(s):  
Jaromi´r Hora´cˇek ◽  
Miloslav Feistauer ◽  
Petr Sva´cˇek
Keyword(s):  
Numerical Simulations ◽  
Degrees Of Freedom ◽  
Stokes Equations ◽  
Mass Matrix ◽  
Vertical Direction ◽  
Navier Stokes ◽  
Computational Domain ◽  
Arbitrary Lagrangian Eulerian ◽  
Navier Stokes Equations ◽  
Classical Flutter

The contribution deals with the numerical simulation of the flutter of an airfoil with three degrees of freedom (3-DOF) for rotation around an elastic axis, oscillation in the vertical direction and rotation of a flap. The finite element (FE) solution of two-dimensional (2-D) incompressible Navier-Stokes equations is coupled with a system of nonlinear ordinary differential equations describing the airfoil vibrations with large amplitudes taking into account the nonlinear mass matrix. The time-dependent computational domain and a moving grid are treated by the Arbitrary Lagrangian-Eulerian (ALE) method and a suitable stabilization of the FE discretization is applied. The developed method was successfully tested by the classical flutter computation of the critical flutter velocity using NASTRAN program considering the linear model of vibrations and the double-lattice aerodynamic theory. The method was applied to the numerical simulations of the post flutter regime in time domain showing Limit Cycle Oscillations (LCO) due to nonlinearities of the flow model and vibrations with large amplitudes. Numerical experiments were performed for the airfoil NACA 0012 respecting the effect of the air space between the flap and the main airfoil.

scholarly journals Effect of Impeller Parameters on the Flow inside the Centrifugal Blower using CFD

2020 ◽  
Vol 8 (6) ◽  
pp. 3977-3980
Keyword(s):  
Numerical Analysis ◽  
Stokes Equations ◽  
Control Volume ◽  
Three Dimensional ◽  
Flow Characteristics ◽  
Navier Stokes ◽  
Computational Domain ◽  
Centrifugal Blower ◽  
Impeller Blade ◽  
Navier Stokes Equations

A numerical analysis is carried out to understand the flow characteristics for different impeller configurations of a single stage centrifugal blower. The volute design is based on constant velocity method. Four different impeller configurations are selected for the analysis. Impeller blade geometry is created with point by point method. Numerical simulation is carried out by CFD software GAMBIT 2.4.6 and FLUENT 6.3.26. GAMBIT work includes geometry definition and grid generation of computational domain. This process includes selection of grid types, grid refinements and defining correct boundary conditions. Processing work is carried out in FLUENT. The viscous Navier-Stokes equations are solved with control volume approach and the k-ε turbulence model. In this three dimensional numerical analysis is carried out with steady flow approach. The rotor and stator interaction is solved by mixing plane approach. Results of simulation are presented in terms of flow parameters, at impeller outlet and various angular positions inside the volute. Also, the contours of flow properties are presented at the outlet plane of fluid domain. Results suggest that for the same configurations of centrifugal blower, as we change geometrical parameter of impeller the flow inside the blower get affected.

scholarly journals Numerical simulation of unsteady flow of a viscous incompressible liquid in flat channels of arbitrary shape of heat exchangers

2020 ◽  
Vol 34 ◽  
pp. 41-55
Author(s):  
Y. Chоvniuk ◽  
V. Kravchuk ◽  
A. Moskvitina ◽  
I. Pefteva
Keyword(s):  
Numerical Simulation ◽  
Fluid Flow ◽  
Unsteady Flow ◽  
Heat Exchangers ◽  
Arbitrary Shape ◽  
Stokes Equations ◽  
Navier Stokes ◽  
Computational Domain ◽  
Two Dimensional ◽  
Navier Stokes Equations

Reasonable development and creation of any device in which there is an interaction between the fluid flow and the elements of the flow parts (for example, heat exchangers, transport and power machines, main pipelines), is impossible without detailed information about the characteristics of the flow, about the forces on the surfaces that are around, about vibroacoustic phenomena, etc. Among the various methods of obtaining information about the characteristics of the flow, about the forces on surfaces that are flown around, about vibroacoustic phenomena, an important role is played by theoretical methods that rely on the equation of hydrodynamics and numerous ways to solve them. In this case, the main efforts are aimed at solving the system of Navier-Stokes equations. In this paper, a general method is described for the numerical solution of the problem of unsteady flow of a viscous incompressible fluid in flat channels of an arbitrary shape of heat exchangers. An effective solution to the problem is achieved by using adaptive networks. The mathematical model of the flow is based on the two-dimensional Navier-Stokes equations in the variables "flow function - vortex" and the Poissonequation for pressure, which are solved on the basis of the finite-difference method. A numerical simulation of the fluid flow in a flat curvilinear elbow is carried out at the Reynolds number Re = 1000. This form reflects the most characteristic features of the flow paths of various hydraulic machines, heat exchangers, hydraulic and pipeline systems. The presentation of the numerical results was carried out on the basis of the VISSIM graphic processing package. One of the main problems (difficulties) in the numerical solution of problems of mathematical physics is the representation of boundary conditions for regions of arbitrary shape. The implementation of various artificial methods that are now used in the approximation of both the curvilinear boundaries themselves and the boundary conditions on them can lead to significant losses in the accuracy of the solution. This is especially evident in problems in which solutions in the boundary region have maximum gradients. An effective method for solving this problem is the use of adapted grids for the computational domain. The essence of this method lies in the fact that such a coordinate system, not necessarily orthogonal, is found in which the boundary lines (surfaces) of the region coincide with the coordinate lines (surfaces). In the flat case, the computational domain is transformed into a rectangular one, and the limit curve is displayed on the sides of the rectangle. In practice, the problem of constructing an adapted mesh is reduced to finding functions that describe the mappings of the canonical (rectangular) region onto the region for which the problem was originally formulated, that is, for the two-dimensional case, the functions x (ξ, η), y (ξ, η) are determined.

scholarly journals A New Mass-Conservative, Two-Dimensional, Semi-Implicit Numerical Scheme for the Solution of the Navier-Stokes Equations in Gravel Bed Rivers with Erodible Fine Sediments

Water ◽  
2020 ◽  
Vol 12 (3) ◽  
pp. 690
Author(s):  
Maurizio Tavelli ◽  
Sebastiano Piccolroaz ◽  
Giulia Stradiotti ◽  
Giuseppe Roberto Pisaturo ◽  
Maurizio Righetti
Keyword(s):  
Numerical Scheme ◽  
Stokes Equations ◽  
Transport Processes ◽  
Sediment Supply ◽  
Navier Stokes ◽  
Computational Domain ◽  
Two Dimensional ◽  
Navier Stokes Equations ◽  
Gravel Bed ◽  
Gravel Bed Rivers

The selective trapping and erosion of fine particles that occur in a gravel bed river have important consequences for its stream ecology, water quality, and overall sediment budgeting. This is particularly relevant in water bodies that experience periodic alternation between sediment supply-limited conditions and high sediment loads, such as downstream from a dam. While experimental efforts have been spent to investigate fine sediment erosion and transport in gravel bed rivers, a comprehensive overview of the leading processes is hampered by the difficulties in performing flow field measurements below the gravel crest level. In this work, a new two-dimensional, semi-implicit numerical scheme for the solution of the Navier-Stokes equations in the presence of deposited and erodible sediment is presented, and tested against analytical solutions and performing numerical tests. The scheme is mass-conservative, computationally efficient, and allows for a fine discretization of the computational domain. Overall, this makes the model suitable to appreciate small-scales phenomena such as inter-grain circulation cells, thus offering a valid alternative to evaluate the shear stress distribution, on which erosion and transport processes depend, compared to traditional experimental approaches. In this work, we present proof-of-concept of the proposed model, while future research will focus on its extension to a three-dimensional and parallelized version, and on its application to real case studies.

Parallel, Multi-Zone, Multi-Block Solver to Study Arterial Branches in Human Vascular System

2001 ◽  
Author(s):  
M. Tadjfar ◽  
R. Himeno
Keyword(s):  
Message Passing ◽  
Incompressible Flows ◽  
Stokes Equations ◽  
Vascular System ◽  
Three Dimensional ◽  
Data Partitioning ◽  
Parallel Execution ◽  
Navier Stokes ◽  
Computational Domain ◽  
Navier Stokes Equations

Abstract The unsteady, three-dimensional, incompressible Navier-Stokes equations are solved numerically to study arterial branches in human vascular system. The solver is capable of dealing with moving boundaries and moving grids. It is designed to handle complex, three-dimensional vascular systems. The computational domain is divided into multiple block subdomains. At each cross section the plane is divided into twelve sub-zones to allow flexibility for handling complex geometries and, if needed, appropriate parallel data partitioning. A second-order in time and third-order upwind finite volume method for solving time-accurate incompressible flows based on pseudo-compressibility and dual time-stepping technique is used. For parallel execution, the flow domain is partitioned. Communication between the subdomains of the flow on Riken’s VPP/700E supercomputer is implemented using MPI message-passing library. The code is capable of running on both shared and/or distributed memory architectures.

Sign in / Sign up

Export Citation Format

Share Document