Xingkai Chi

CFD Analysis of the Aerodynamics of a Business-Jet Airfoil with Leading-Edge Ice Accretion

For rime ice - where the ice buildup has only rough and jagged surfaces but no protruding horns - this study shows two dimensional CFD analysis based on the one-equation Spalart-Almaras (S-A) turbulence model to predict accurately the lift, drag, and pressure coefficients up to near the stall angle. For glaze ice - where the ice buildup has two or more protruding horns near the airfoils leading edge - CFD predictions were much less satisfactory because of the large separated region produced by the horns even at zero angle of attack. This CFD study, based on the WIND and the Fluent codes, assesses the following turbulence models by comparing predictions with available experimental data: S-A, standard k-epsilon, shear-stress transport, v(exp 2)-f, and differential Reynolds stress.

Estimating Grid-Induced Errors in Navier-Stokes Solutions by Euler Discrete-Error-Transport Equations

This paper presents and evaluates a method for estimating grid-induced errors in CFD solutions that recognizes error at one location in the flow domain may not be generated there, but rather generated elsewhere and then transported there. This paper derives a system of discrete error-transport equations (DETEs) to compute the evolution of grid-induced errors in finite-volume solutions of the Euler equations for compressible flows in two dimensions. These DETEs are then used to estimate grid-induced errors of Navier-Stokes solutions obtained by using the Fluent code on the basis that error transport is mostly by convection and not by diffusion. Results for a test problem involving compressible low Mach number flow over an iced airfoil show that if the residuals in the DETEs are modeled accurately, then the DETEs can predict grid-induced errors accurately.

Direct Simulation of Surface Roughness Effects with a RANS and DES Approach on Viscous Adaptive Cartesian Grids

The main objective of this research is to directly compute the skin friction (cf) and heat transfer (St) coefficients on real rough surfaces using a state-of-the-art unstructured adaptive grid-based finite volume method. Recent experiments with real roughness panels by Bons are computationally simulated in this study. Computational results are compared with experimental data to assess the simulation accuracy. A RANS (Reynolds-Averaged Navier-Stokes) approach based on the Spalart-Allmaras turbulence model and a DES (Detached Eddy Simulation) approach are employed for the computations, and grid refinement studies are conducted to assess the effects of grid resolution. In two cases with rough surfaces, the RANS approach is capable of accurately predicting cf (within 3.5%) while under-predicting St by 8-15%. The DES approach was able to predict cf and St for smooth flat panels but failed in the cases with real roughness. The cause will be further investigated.

Bulk Temperature, Heat-Transfer Coefficient, and Nusselt Number - Revisited

Convective heat transfer on surfaces is often presented in the form of the heat-transfer coefficient (h) or nondimensionally in the form of the Nusselt number (Nu). For internal flows, h and Nu depend on the bulk temperature in addition to the geometry and the nature of the flow. However, the bulk temperature is clearly defined only for a simple duct without flow separation. Also, even when it is well defined, it is hard to measure experimentally so that it is often approximated when experimentalists measure h. In this study, computational fluid dynamics (CFD) is used to examine several approximations of the bulk temperature that are commonly used in experimental measurements and how these approximations affect the measured heat-transfer coefficient and Nusselt number. The effects of compressibility and temperature-dependent properties on heat-transfer and Nusselt are also examined. The test problem used is flow and heat transfer in a straight duct with a circular cross section under laminar and turbulent conditions and under compressible and incompressible flow conditions with constant and temperature-dependent properties. This test problem was selected because exact solutions and well-established correlations exist for the incompressible-flow cases to validate the CFD and to assess the errors created by the approximations to the bulk temperature. Nomenclature

Effects of Pin-Fin Height on Flow and Heat Transfer in a Rectangular Duct

CFD simulations were performed to study the flow and heat transfer in a rectangular duct (Wd × Hd , where Wd /Hd = 3) with a staggered array of circular pin fins (D = Hd /4) mounted on the two opposite walls separated by Hd . For this array of pin fins, five different pin-fin height (H) combinations were examined, and they are (1) H = Hd = 4D (i.e., all pin fins extended from wall to wall), (2) H = 3D on both walls, (3) H = 2D on both walls, (4) H = 4D on one wall and H = 2D on the opposite wall, and (5) H = 3D on one wall and H = 2D on the opposite wall. The H values studied give H/D values of 2, 3, and 4 and C/D values of 2, 1, and 0, where C is the distance between the pin-fin tip and the opposite wall. For all cases, the duct wall and pin-fin surface temperatures were maintained at Tw = 313.15 K; the temperature and the speed of the air at the duct inlet were uniform at Tinlet = 343.15 K and U = 8.24 m/s; the pressure at the duct exit was fixed at Pb = 1 atm; and the Reynolds number based on the duct hydraulic diameter and duct inlet conditions was Re = 15,000. This CFD study is based on 3-D steady RANS, where the ensemble averaged continuity, compressible Navier-Stokes, and energy equations are closed by the thermally perfect equation of state and the two-equation realizable k-e turbulence model with wall functions and with the low-Reynolds number model of Chen and Patel in the near-wall region. The usefulness of this CFD study was assessed by comparing predicted heat-transfer coefficient and friction factor with available experimental data. Results are presented to show how the flow induced by arrays of pin fins of different heights affects temperature distribution, surface heat transfer, and pressure loss.Copyright

Q3D-Wing: Validation of a Modern Lifting-Line Method for Application to Clean and Iced Wings

Though computational fluid dynamics (CFD) based on the Navier-Stokes equations are useful in analyzing the aerodynamics of finite wings, they are impractical for studying wings that have ice formed on them because of the enormous number of grid points/cells needed to resolve the ice geometry. In this study, a practical engineering method for computing the aerodynamics of clean and iced wings, referred to as Q3DWing, is presented and evaluated. With this method, two-dimensional (2D) CFD analyses are generated for a number of airfoil sections as a function of angle of attack, and these 2D CFD results are then connected to form the wing by the lifting-line theory to predict 3D clean or iced wings aerodynamics. Q3D-Wing is evaluated by comparing predictions with those from 3D CFD simulations as well as by wings for which there are experimental data for the 2D airfoil sections and 3D wings formed by those 2D airfoil sections. Results obtained show Q3D-Wing to provide good agreement for clean and iced wings.

Flow and Heat Transfer in the Tip-Turn Region of a U-Duct Under Rotating and Non-Rotating Conditions

CFD simulations were performed to study the flow and heat transfer in a U-duct, relevant to internal cooling of the first-stage turbine component in electric-power-generation, gas-turbine engines. Parameters studied include (1) two aspect ratios of the duct cross section, i.e. H/W = 1 and H/W = 0.25; (2) smooth duct and duct lined with pin fins of height H arranged in a staggered fashion; and (3) two rotational speeds: 0 rpm and 3,600 rpm. In all cases, the wall temperature is 1173 K; the coolant temperature at the U-duct inlet is 623 K; and the back pressure at the exit of the U-duct is 25.17 atm. The Reynolds numbers studied are 150,000 for the duct with the 4-to-1 aspect ratio, and 150,000 and 375,000 for the duct with the 1-to-1 aspect ratio. When there is rotation at 3,600 rpm, the rotational numbers corresponding to these Reynolds numbers and duct aspect ratios are 0.592, 1.64, and 4.11, respectively. Result is presented to show the nature of the flow, the temperature distribution, and the surface heat transfer with focus on the flow and heat transfer in the tip-turn region as a function of the parameters investigated. This computational study is based on 3-D steady RANS. The ensemble-averaged continuity, compressible Navier-Stokes, and energy equations were closed by the thermally perfect equation of state with temperature-dependent gas properties and the two-equation realizeable k-e turbulence model with and without wall functions.Copyright