Numerical Design and Study of a MEMS-Based Micro Turbine

2005 ◽  
Author(s):  
Tatsuo Onishi ◽  
Ste´phane Burguburu ◽  
Olivier Dessornes ◽  
Yves Ribaud
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Skin Friction ◽  
Large Scale ◽  
Low Reynolds Number ◽  
Turbine Blades ◽  
Three Dimensional ◽  
Navier Stokes ◽  
Low Reynolds ◽  
Micro Turbine

A full three dimensional Navier-Stokes solver elsA developed by ONERA is used to design and study the aerothermodynamics of a MEMS-based micro turbine. This work is performed in the framework of micro turbomachinery project at ONERA. A few millimeter scale micro turbine is operated in a low Reynolds number regime (Re = 5,000∼50,000), which implies a more important influence of skin friction and heat transfer than the conventional large-scale gas turbine. The 2D geometry constraints due to the limitation of fabrication technology also distinguish the aerothermodynamic characteristics of a micro turbine from that of conventional turbomachinery. Thus, for the foundation of aerothermodynamic design of micro turbomachinery, understanding of low Reynolds number effects on the performance is required and then the design of the turbine geometry can be optimized. In this study, aero-thermodynamic effects at low Reynolds number and different stator/rotor configurations are examined with a prescribed wall temperature. Losses due to heat transfer to walls and skin friction are estimated and their effects on the operating performance are discussed. Power delivery to turbine blades is checked and found satisfactory to give the objective design value of more than 100W. The effects of turbine exhaust geometry and the number of blades on turbine performance are also discussed.

Numerical Simulation of Turbine Blade Boundary Layer and Heat Transfer and Assessment of Turbulence Models

1997 ◽  
Vol 119 (4) ◽  
pp. 794-801 ◽  
Author(s):  
J. Luo ◽  
B. Lakshminarayana
Keyword(s):  
Heat Transfer ◽  
Boundary Layer ◽  
Reynolds Number ◽  
Low Reynolds Number ◽  
Turbulence Models ◽  
Navier Stokes ◽  
Bypass Transition ◽  
Boundary Layer Development ◽  
Low Reynolds ◽  
Layer Development

The boundary layer development and convective heat transfer on transonic turbine nozzle vanes are investigated using a compressible Navier–Stokes code with three low-Reynolds-number k–ε models. The mean-flow and turbulence transport equations are integrated by a four-stage Runge–Kutta scheme. Numerical predictions are compared with the experimental data acquired at Allison Engine Company. An assessment of the performance of various turbulence models is carried out. The two modes of transition, bypass transition and separation-induced transition, are studied comparatively. Effects of blade surface pressure gradients, free-stream turbulence level, and Reynolds number on the blade boundary layer development, particularly transition onset, are examined. Predictions from a parabolic boundary layer code are included for comparison with those from the elliptic Navier–Stokes code. The present study indicates that the turbine external heat transfer, under real engine conditions, can be predicted well by the Navier–Stokes procedure with the low-Reynolds-number k–ε models employed.

Flow Characteristics in a Three-Dimensional Rectangular Channel With a Pair of Ribs Placed Symmetrically at the Channel Walls

2000 ◽  
Author(s):  
M. Singh ◽  
P. K. Panigrahi ◽  
G. Biswas
Keyword(s):  
Heat Transfer ◽  
Large Scale ◽  
Numerical Study ◽  
Rectangular Channel ◽  
Turbine Blades ◽  
Three Dimensional ◽  
Flow Characteristics ◽  
Navier Stokes ◽  
Complex Flow ◽  
Energy Equations

Abstract A numerical study of rib augmented cooling of turbine blades is reported in this paper. The time-dependent velocity field around a pair of symmetrically placed ribs on the walls of a three-dimensional rectangular channel was studied by use of a modified version of Marker-And-Cell algorithm to solve the unsteady incompressible Navier-Stokes and energy equations. The flow structures are presented with the help of instantaneous velocity vector and vorticity fields, FFT and time averaged and rms values of components of velocity. The spanwise averaged Nusselt number is found to increase at the locations of reattachment. The numerical results are compared with available numerical and experimental results. The presence of ribs leads to complex flow fields with regions of flow separation before and after the ribs. Each interruption in the flow field due to the surface mounted rib enables the velocity distribution to be more homogeneous and a new boundary layer starts developing downstream of the rib. The heat transfer is primarily enhanced due to the decrease in the thermal resistance owing to the thinner boundary layers on the interrupted surfaces. Another reason for heat transfer enhancement can be attributed to the mixing induced by large-scale structures present downstream of the separation point.

Heat Transfer Investigations in Multiple Impinging Jets at Low Reynolds Number

2010 ◽  
Author(s):  
Arvind G. Rao ◽  
Myra Kitron-Belinkov ◽  
Vladimir Krapp ◽  
Yeshayahou Levy
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Low Reynolds Number ◽  
Turbine Blades ◽  
Jet Impingement ◽  
High Heat ◽  
Geometrical Parameters ◽  
Transitional Region ◽  
High Heat Transfer ◽  
Low Reynolds

Jet impingement is a well established cooling methodology used for cooling turbine blades in gas turbine engines. Jet impingement results in high heat transfer coefficients as compared to other conventional modes of single phase heat transfer. Most of the research in jet impingement has been confined to high Reynolds number regime. In order to increase the applicability of this technique to non conventional applications like in a low pressure micro turbine combustors or turbine blades, the behavior of such systems in the low Reynolds number regime should be understood. The present paper is a continuation of earlier investigations on the heat transfer behavior of a large jet impingement array in the low Reynolds number regime, especially in the laminar and transitional region. More experiments have been conducted with different geometrical parameters of the array to analyze the effect of these parameters on the average heat transfer coefficient. Numerical simulations with existing CFD tools were carried out in order to understand the fluid mechanics inside such a complex system. The CFD model was validated with the experiments. Different turbulence models were used and it was found that the SST-k-ω model was the best for modeling jet impingement phenomena. It is anticipated that the results obtained from the present exercise will give better insights in optimizing the design of multiple jet impingement cooling systems for high heat density applications.

Validation of Numerical Analysis to Estimate Airfoil Aerodynamic Characteristics at Low Reynolds Number Region

2015 ◽  
Author(s):  
Donghwi Lee ◽  
Taku Nonomura ◽  
Akira Oyama ◽  
Kozo Fujii
Keyword(s):  
Reynolds Number ◽  
Numerical Method ◽  
Low Reynolds Number ◽  
Lift Coefficient ◽  
Three Dimensional ◽  
Aerodynamic Characteristics ◽  
Navier Stokes ◽  
Two Dimensional ◽  
Eddy Simulation ◽  
Low Reynolds

In this study, two-dimensional laminar simulation (2-D Lam), two-dimensional Reynolds Averaged Navier-Stokes simulation with the Spalart-Allmaras turbulence model (2-D RANS(SA)), and implicit three-dimensional large-eddy simulation (3-D LES) are performed for NACA0012, NACA0006, and Ishii airfoils at Rec = 3.0 × 104. The relation between a predictability of airfoil aerodynamic characteristics and a dependence of airfoil geometry shape of each numerical method is evaluated at the low Reynolds number. Although little discrepancy is observed for the lift coefficient predictability, significant differences are presented in terms of the separation and reattachment points predictability depending on the numerical methods. The 2-D Lam simulation can predict the lift coefficients as well as the separation and reattachment points qualitatively as similar to the 3-D LES results except for the high angle of attack which is accompanied by the massive separation. The 2-D RANS(SA), the weak nonlinearity and stall phenomena for the lift coefficients are observed. A good predictability of the separation point are shown, however, it cannot be estimated the reattachment points due to the trend to predict widely for the separation region. The predictabilities of each numerical method appear regardless of the airfoil shapes.

3D Numerical Analysis of Heat Transfer in a Low Reynolds Number Microturbine Cascade

2012 ◽  
Author(s):  
M. Omri ◽  
S. Moreau ◽  
L. G. Fréchette
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Waste Heat ◽  
Low Reynolds Number ◽  
Three Dimensional ◽  
Suction Side ◽  
Tip Clearance ◽  
Tip Vortex ◽  
Distributed Power Generation ◽  
Low Reynolds

This paper presents the conjugate heat transfer in a submillimeter scale microturbine characterized by laminar yet highly three-dimensional flows. Such a miniature turbine is part of a MEMS (microelectromechanical system) power plant-on-a-chip currently under development for distributed power generation from waste heat. Adiabatic subsonic flows in the turbine have previously been studied numerically and are characterized by low Reynolds number laminar flow (Re < 2500) but with complex vortical structures. The present work addresses the influence of these flow structures on heat transfer, including the effect of the horseshoe and tip vortices. Calculations were done for tip clearance gaps equal to 0%, 5% and 10% blade height. Three different scenarios were considered: adiabatic walls, the hub and casing temperature of 573K or the hub at 573K and the casing at 450K, for incoming flow at 600K. The heat transfer is more variable in the suction side since dominant vortices are adjacent to this blade side. The heat flux even changes its sign where the vortices begin to separate from the suction side, indicating that gas cooled in the hub and casing boundary layers is transported on the blades by the horseshoe vortices. The tip vortex prevents the top passage vortex from interacting with the suction side, which eliminates the negative heat transfer in this region. Due to the dominant vortices, the Nusselt number is found to be a function of the thermal boundary conditions and cannot be predicted with traditional boundary layer correlations.

Analysis of transitional flow and heat transfer over turbine blades: Algebraic versus low-Reynolds-number turbulence model

2001 ◽  
Vol 215 (9) ◽  
pp. 1003-1018 ◽  
Author(s):  
S Sarkar
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Low Reynolds Number ◽  
Turbine Blades ◽  
Transitional Flow ◽  
Flow And Heat Transfer ◽  
Turbulent Transition ◽  
Wide Range ◽  
Low Reynolds ◽  
Laminar Turbulent Transition

The numerical simulation of flow and heat transfer over turbine blades involving laminar-turbulent transition is presented. The predicted results are compared with the experimental surface heat transfer and pressure distributions for two transonic turbine blades over a wide range of flow conditions. The time-dependent, mass-averaged Navier-Stokes equations are solved by an explicit four-stage Runge-Kutta scheme in the finite volume formulation. Local time stepping, variable-coefficient implicit residual smoothing and a full multigrid method have been implemented to accelerate the steady state calculation. The turbulence is simulated by the algebraic Baldwin-Lomax model together with an explicitly imposed model for transition. For comparison, the low-Reynolds-number version of the two-equation ( k-∊) model of Chien is also used. The modified Baldwin-Lomax model performs well in predicting the onset of laminar-turbulent transition, whereas the Chien model shows a tendency to mimic the transition early and over a shorter distance.

Sign in / Sign up

Export Citation Format

Share Document