Introduction to Computational I t d ti t C t ti l Fluid Dynamics (CFD

)
Tao Xing and Fred Stern

Outline
1. 2. 2 3. 4. 5. 6. 7. 8. What, why and where of CFD? Modeling M d li Numerical methods Types of CFD codes CFD Educational Interface CFD Process Example of CFD Process 58:160 CFD Labs

2

What is CFD?

CFD is the simulation of fluids engineering systems using modeling (mathematical physical problem formulation) and numerical methods (discretization methods, solvers, numerical i l th d (di ti ti th d l i l parameters, and grid generations, etc.) Historically only Analytical Fluid Dynamics (AFD) and Experimental Fluid Dynamics (EFD). E i t l Fl id D i (EFD) CFD made possible by the advent of digital computer and advancing with improvements of computer resources (500 flops, 1947 20 teraflops, 2003)

• •

3

Why use CFD? y
• Analysis and Design
More cost effective and more rapid than EFD CFD provides high-fidelity database for diagnosing flow field

2. Si l i 2 Simulation of physical fluid phenomena that are f h i l fl id h h difficult for experiments
Full scale simulations (e.g., ships and airplanes) Environmental effects (wind weather etc ) (wind, weather, etc.) Hazards (e.g., explosions, radiation, pollution) Physics (e.g., planetary boundary layer, stellar evolution) )

• Knowledge and exploration of flow physics

4

Where is CFD used?
• Where is CFD used? • Aerospace • Automotive • Biomedical
• Chemical • • • • • •
Processing HVAC Hydraulics Marine Oil & Gas Power Generation Sports
Automotive Aerospace

Biomedical
F18 Store Separation

Temperature and natural convection currents in the eye ti t i th following laser heating.

5

Chemical Processing Chemical Processing • HVAC • H d li Hydraulics • Marine • Oil & Gas • Power G Generation • Sports HVAC Hydraulics Streamlines for workstation ventilation 6 .Where is CFD used? • Where is CFD used? • • • • Aerospacee Automotive Biomedical Polymerization reactor vessel .prediction of flow separation and residence time effects.

Where is CFD used? Marine (movie) Sports • Where is CFD used? • • • • • • • • • • Aerospace Automotive Biomedical Chemical Processing HVAC Hydraulics y Marine Oil & Gas Power Generation Sports Oil & Gas Flow of lubricating mud over drill bit Power Generation Flow around cooling towers 7 .

conditions a d initial co d t o s • Modeling includes: 1. Flow conditions 5. Geometry and domain 2. 2 Coordinates 3. Initial and boundary conditions 6. Selection of models for different applications 8 . Governing equations 4.Modeling • Modeling is the mathematical physics problem formulation in terms of a continuous initial boundary value problem (IBVP) • IBVP is in the form of Partial Differential Equations (PDEs) with appropriate boundary co d t o s and t a conditions.

appropriately chosen for a better resolution of the geometry (e. cylindrical for circular pipe). system(r θ. ACIS. cylindrical system (r.g. or IGES. • The three coordinates: Cartesian system (x. θ z) and spherical system(r. STEP. θ Φ) should be (r θ. 9 .y. etc.Modeling (geometry and domain) g (g y ) • Simple geometries can be easily created by few geometric parameters (e.g. airfoil) into commercial software • Domain: size and shape • Typical approaches • Geometry approximation • CAD/CAE integration: use of industry standards such as Parasolid.z). z). circular pipe) • Complex geometries must be created by the p p g y partial differential equations or importing the database of the geometry(e.g.

φ) φ y θ r General Curvilinear Coordinates General orthogonal Coordinates 10 .z) z Cylindrical (r.θ.y.θ.Modeling (coordinates) z Cartesian (x.z) z y x x θ r y x z Spherical (r.

• Modeling (governing equations) Navier-Stokes N i St k equations (3D i C t i coordinates) ti in Cartesian di t ) ⎡ ∂ 2u ∂ 2u ∂ 2u ⎤ ˆ ∂p ∂u ∂u ∂u ∂u ρ + μ⎢ 2 + 2 + 2 ⎥ =− + ρw + ρv + ρu ∂z ⎦ ∂y ∂x ∂z ∂y ∂x ∂t ⎣ ∂x ⎡ ∂ 2v ∂ 2v ∂ 2v ⎤ ˆ ∂p ∂v ∂v ∂v ∂v ρ + μ⎢ 2 + 2 + 2 ⎥ =− + ρw + ρv + ρu ∂z ⎦ ∂y ∂y ∂z ∂y ∂x ∂t ⎣ ∂x ⎡ ∂2w ∂2w ∂2w ⎤ ˆ ∂w ∂w ∂w ∂w ∂p ρ + ρu + ρv + ρw = − + μ⎢ 2 + 2 + 2 ⎥ ∂t ∂x ∂y ∂z ∂z ∂y ∂z ⎦ ⎣ ∂x Local acceleration Convection Piezometric pressure gradient Viscous terms ∂ρ ∂ ( ρ u ) ∂ ( ρ v ) ∂ ( ρ w ) + + + = 0 Continuity equation ∂t ∂x ∂y ∂z p = ρRT p − p D 2R 3 DR 2 + ( R ) = v Dt 2 ρL 2 Dt Equation of state Rayleigh Equation 11 .

compressible (Ma) • Single. γ. Gr. laminar (Re) • Incompressible vs. multi-phase (Ca) l l h ( ) • Thermal/density effects (Pr. CFD • Viscous vs. Ec) • Free-surface flow (Fr) and surface tension (We) Free surface • Chemical reactions and combustion (Pe.Modeling (flow conditions) can be distinguished into different categories using different criteria • Based on the physics of the fluids phenomena. Da) • etc… 12 . inviscid (Re) • External flow or internal flow (wall bounded or not) ( ) • Turbulent vs.vs.

steady/unsteady flows) • ICs should not affect final results and only y affect convergence path. number of iterations (steady) or time steps (unsteady) need to reach converged solutions solutions. CFD codes are usually run in the steady mode for a few iterations for getting a better initial conditions 13 . • More reasonable guess can speed up the convergence • For complicated unsteady flow problems. i.e.Modeling (initial conditions) • Initial conditions (ICS.

No-slip walls: u=0. etc.du/dr=0 Axisymmetric 14 . mass flow rate. pressure etc ) outlet (constant pressure. numerical beach. and nonreflecting (for compressible flows. inlet (velocity inlet. etc. constant pressure.v=0 r Outlet. velocity pressure convective. p=c Periodic boundary condition in spanwise direction of an airfoil o x v=0. such as acoustics). zero-gradient). •Boundary conditions: No-slip or slip-free on walls.u=c.v=0 Inlet .). dp/dr=0.Modeling(boundary conditions) g( y ) periodic.

Gr. γ. multi-phase (Ca. equation of state) p p ( . conservation of energy) • Free-surface flow (Fr. laminar (Re. Ec. bubble dynamic model) • Chemical reactions and combustion (Chemical reaction model) • etc… 15 . two-fluid model) • Th Thermal/density effects and energy equation l/d i ff d i (Pr. compressible (Ma. level-set & surface tracking model) and ( . g ) surface tension (We.Modeling (selection of models) • CFD codes typically designed for solving certain fluid phenomenon by applying different models • Vi Viscous vs. q ) • Single. Turbulent models) • Incompressible vs. i i id inviscid (Re) (R ) • Turbulent vs. cavitation model.vs.

efficient inside BL but excessive • LES: accurate in separation region and unaffordable for resolving BL • Free-surface models: • DES: RANS inside BL. • Surface-tracking method: mesh moving to capture free surface. Surface tracking limited to small and medium wave slopes diffusion in the separated region. studying steep or breaking waves. LES in separated regions. • Single/two phase level-set method: mesh fixed and level-set function used to capture the gas/liquid interface capable of interface. 16 . but too expensive for turbulent flows • RANS: predict mean flow structures.Modeling (Turbulence and free surface models) • Turbulent flows at high Re usually involve both large and small scale • Turbulent models: vortical structures and very thin turbulent boundary layer (BL) near the wall • DNS: most accurately solve NS equations.

Iso-surface of Q criterion (0. contour of vorticity for turbulent flow around NACA12 with angle of attack 60 degrees DES. Wigley Hull pitching and heaving 17 . Re=105.4) for turbulent flow around NACA12 with angle of attack 60 degrees URANS. Re=105.Examples of modeling (Turbulence and free surface models) ) URANS.

Discretization methods Solvers and numerical parameters S l d i l t Grid generation and transformation High Performance Computation (HPC) and postprocessing 18 . Assemble the system of algebraic equations and solve the system to get approximate solutions • Numerical methods include: 1. 2 3.Numerical methods • The continuous Initial Boundary Value Problems (IBVPs) are discretized into algebraic equations using numerical methods. 4. 2.

etc. Runge-Kutta. usually for regular grid) and finite volumes and finite element methods (usually for irregular meshes) ( s all fo i eg la • Each type of methods above yields the same solution if the grid is fine enough. central differences schemes.Discretization methods • Finite difference methods (straightforward to apply.) or implicit method (e. some methods are more suitable to some cases than others • Finite difference methods for spatial derivatives with different order of accuracies can be derived using Taylor expansions. etc. but more likely unstable due to less numerical dissipation • Temporal derivatives can be integrated either by the explicit method (Euler. However. • Higher order numerical methods usually predict higher order of accuracy for CFD. Beam-Warming method) 19 . such as 2nd order upwind scheme.g.

Discretization methods (Cont’d) ( ) • Explicit methods can be easily applied but yield conditionally stable Finite Different Equations (FDEs). such as multi-phase flows. • Selection of discretization methods should consider efficiency. • Pre-conditioning method is used when the matrix of the linear algebraic system is ill-posed. accuracy and special requirements. flows with a broad range of Mach numbers. such as shock wave tracking. are unconditionally stable. higher-order temporal discretization is used y g p when the spatial discretization is also of higher order. • Stability: A discretization method is said to be stable if it does not magnify the errors that appear in the course of numerical solution process. which are restricted b the time step Implicit methods hich a e est icted by step. but need efforts on efficiency. 20 . etc. • Usually.

m+1) ∂u ∂v =0 + ∂x ∂y (L-1. l ∂u vm l l ⎡um +1 − um ⎤ FD Sign( v l )<0 = v ⎦ m ∂y Δy ⎣ l vm l l ⎡um − um −1 ⎤ BD Sign( l )>0 = ⎦ Δy ⎣ vm 1st order upwind scheme.e. theoretical order of accuracy Pkest= 1 21 .m-1) L-1 L x ∂ 2u μ l l l μ 2 = 2 ⎡um +1 − 2um + um −1 ⎤ ⎦ ∂y Δy ⎣ 2nd order central difference i.e..m) (L.m) m 1 m=1 m=0 y m=MM+1 m=MM ∂u ∂u ∂ ⎛ p⎞ ∂ 2u +v = − ⎜ ⎟+μ 2 u ∂x ⎝ e ⎠ ∂x ∂y ∂y l ∂u um l l ⎡um − um−1 ⎤ u = ⎦ ∂x Δx ⎣ (L. i. theoretical order of accuracy Pkest= 2.Discretization methods (example) • 2D incompressible laminar flow boundary layer (L..

22 .Discretization methods (example) 3 1 B1 ⎡ B2 ⎤ − FD l l ⎢ ul ⎤ l ⎡ μ ⎤ l 2μ ⎥ l ⎡ μ vm vm Δy l m − 2 ⎥ um + ⎢ 2 + FD ⎥ um +1 + ⎢ 2 − BD ⎥ um −1 ⎢ + vm 1 Δy ⎢ Δx ⎣ Δy ⎦ ⎣ Δy Δy ⎦ BD Δy ⎥ ⎢ ⎥ Δy ⎣ ⎦ B l l l l B1um −1 + B2um + B3um +1 = B4um−1 − l ⎡ ∂ ⎛ p ⎞ ⎤ S l it using l −1 i ⎢ B4u1 − ⎜ ⎟ ⎥ Solve l ∂x ⎝ e ⎠1 ⎥ Thomas algorithm ⎡ B2 B3 0 0 0 0 0 0 ⎤ ⎡ u1 ⎤ ⎢ ⎢ ⎥ ⎢B B B ⎥ 0 0 0 0 0⎥ ⎢ • ⎥ ⎢ • 1 2 3 ⎢ ⎥ ⎢ ⎥ ⎢ ⎥×⎢ • ⎥ = ⎢ • • • • • ⎥ ⎥ ⎢ ⎢ ⎥ ⎢ ⎥ 0 0 0 0 0 B1 B2 B3 ⎥ ⎢ • ⎥ • ⎢ ⎢ ⎥ l l ⎢ 0 0 0 0 0 0 B1 B2 ⎥ ⎢umm ⎥ ⎢ ⎣ ⎦ ⎣ ∂ ⎛ p⎞ ⎥ ⎦ l −1 ⎢ B4umm − ⎜ ⎟ ⎥ ∂x ⎝ e ⎠ mm ⎥ ⎢ ⎣ ⎦ To be stable. Matrix has to be ∂ l p / e )m ( ∂x l um l −1 ∂ = um − ( p / e)lm Δx ∂x B4 Diagonally dominant.

Gauss-Seidel method method Gauss Seidel method. • Under relaxation factor convergence limit etc factor.Solvers and numerical parameters p • Solvers include: tridiagonal. SOR method) • Numerical parameters need to be specified to control the calculation. Gauss elimination. • Solvers can be either direct (Cramer’s rule. limit. etc solvers etc. LU decomposition) or iterative (Jacobi method. solution-adaptive solver. etc. PETSC solver. multi-grid solvers. pentadiagonal solvers. • Different numerical schemes • Monitor residuals (change of results between iterations) it ti ) • Number of iterations for steady flow or number of time steps for unsteady flow • Single/double precisions 23 .

or regions interested (FLUENT) structured unstructured 24 .Numerical methods (grid generation) • Grids can either be structured (hexahedral) or unstructured (tetrahedral). Depends upon type of discretization scheme and application • Scheme Finite differences: structured Finite l Fi it volume or finite element: fi it l t structured or unstructured • Application Thin boundary layers best resolved with highly-stretched structured grids Unstructured grids useful for complex geometries Unstructured grids permit automatic adaptive refinement based on the pressure gradient.

ζ) domains.η. (ξ η ζ) domains important for body-fitted grids. The partial ∂f ∂f ∂ξ ∂f ∂η ∂f ∂f = + = ξy +ηy derivatives at these two domains have the ∂y ∂ξ ∂y ∂η ∂y ∂ξ ∂η relationship ( as an example) p (2D p ) •Transformation between physical (x.z) 25 .y Numerical methods (grid transformation) η Transform o Physical domain x o Computational domain ∂f ∂f ∂ξ ∂f ∂η ∂f ∂f = + = ξx + ηx ∂x ∂ξ ∂x ∂η ∂x ∂ξ ∂η ξ and computational (ξ.y.

CFD codes need to be developed using the Massage Passing Interface (MPI) Standard to transfer data between different blocks.). • Post-processing: 1 Visualize the CFD results (contour.g. pathlines. and 2. CFD UA: verification and validation using EFD data (more details later) • Post processing usually through using commercial software Post-processing 26 . and iso-surface in 3D. etc.• CFD computations ( t ti (e. streak lines. 3D unsteady fl t d flows) are usually ) ll High performance computing and postprocessing very expensive which requires parallel high performance supercomputers (e. IBM 690) with the use of multi-block technique. • As required by the multi-block technique.g. (contour velocity vectors. Post processing: 1. streamlines.

FieldView) CFDSHIPIOWA 27 . Gridgen. Tecplot. etc.g. CFX/AEA. Gambit) and flow visualization software (e. etc.Types of CFD codes yp • Commercial CFD code: FLUENT.) • Other CFD software includes the Grid generation software (e. • Research CFD code: CFDSHIP IOWA CFDSHIP-IOWA • Public domain software (PHI3D.g. StarCD. and WinpipeD. HYDRO. CFDRC.

EFD. Validation using EFD Lab 2: Airfoil Flow 1. Effect of slant angle and comparison with LES. Meshing and iterative convergence 2. Boundary layer separation 3. Iterative error 4. “C” and “O” Meshes 4. Axial velocity profile 4. EFD and RANS RANS. Boundary layer separation 3. Streamlines 5. Lab4: Ahmed car 1. 6 Verification using AFD 7. Boundary conditions 3. Grid error 5. Effect of turbulence models 6.CFD Educational Interface Lab1: Pipe Flow 1. Effect of expansion angle and comparison with LES. EFD. Boundary conditions 2. and RANS. Developing length of laminar and turbulent pipe flows. Verification and Validation using EFD Lab3: Diffuser 1. Streamlines 5. 28 . 6. Effect of angle of attack/turbulent models on flow field 5. Meshing and iterative convergence 2. Definition of “CFD Process” 2. Axial velocity profile 4. Effect of order of accuracy on verification results 3. Effect of grid generation topology.

) • Once purposes and CFD codes chosen. Solve 5. combustion. multi-phase li i ( i b i li h flows. • Depend on the specific purpose and flow conditions of the problem. Reports 6. Geometry 2.CFD process p • Purposes of CFD codes will be different for different applications: investigation of bubble-fluid interactions for bubbly flows. Physics 3. study of wave induced massively separated flows for free-surface. “CFD process” is the steps to set up the IBVP problem and run the code: 1. etc. marines. Post processing 29 . Mesh 4. etc. different CFD codes can be chosen for different applications (aerospace.

etc) Geometry Parameters Compressible ON/OFF Structured (automatic/ manual) Iterations/ Steps XY Plot Vectors Domain Shape and Size Flow properties Convergent Limit Verification Streamlines Viscous Model Precisions (single/ double) Validation Boundary Conditions Numerical Scheme Initial Conditions 30 . shear stress.CFD Process Geometry Physics Mesh Solve Reports PostProcessing Contours Select Geometry Heat Transfer ON/OFF Unstructured (automatic/ manual) Steady/ Unsteady Forces Report (lift/drag.

g. commercial software (e. geometry is usually created using commercial software (either separated from the commercial code itself. Gridgen) is used. or combined together. like Gambit.Geometry y • • • • Selection of an appropriate coordinate Determine the domain size and shape p Any simplifications needed? What kinds of shapes needed to be used to best resolve the geometry? (lines.) • For commercial code. etc. 31 . like FlowLab) • For research code. ovals. circular.

etc. or turbulent. 2. y. Selection of models: different models usually fixed by codes. laminar. and thermal conductivity. Fluid properties: density. etc. viscosity. user needs specify them for different applications. opt o s for use to c oose ed options o user choose Initial and Boundary Conditions: not fixed by codes. viscous. 3. Flow conditions and properties usually presented in dimensional form in industrial commercial CFD software. Flow conditions: inviscid. 32 • • . whereas in non.Physics • Flow conditions and fluid properties 1. dimensional variables for research codes.

conformal mapping. Gambit. together with the boundary conditions need to be exported from commercial software in a certain format that can be recognized by the research CFD code or other commercial CFD software. 33 . etc.. PDE based. such as the grid refinement inside the wall boundary layer • Mesh can be gene ated b either commercial codes generated by eithe comme cial (Gridgen. Re).) • The mesh.) or research code (using algebraic vs. etc.g.Mesh • Meshes should be well designed to resolve p p p important flow features which are dependent upon flow condition parameters (e.

drag forces) 34 .g. y. p integral quantities (lift. pressure and y.Solve • Setup appropriate numerical parameters • Ch Choose appropriate Solvers i t S l • Solution procedure (e. incompressible flows) Solve the momentum pressure Poisson momentum. turbulence intensity. such as velocity. equations and get flow field quantities.

etc. friction factor (pipe flow). friction factor distribution (pipe flow). pressure and temperature.Reports p • Reports saved the time history of the residuals of the velocity. etc. • Report the integral quantities. such as total pressure drop. g ( ). • AFD or EFD data can be imported and put on top of the XY plots for validation 35 . lift and drag coefficients (airfoil flow). pressure coefficient flow) distribution (airfoil flow). • XY plots could present the centerline velocity/pressure distribution.

Post-processing • Analysis and visualization • Calculation of derived variables Vorticity Wall shear stress • Calculation of integral parameters: forces. moments • Visualization (usually with commercial software) Simple Si l 2D contours t 3D contour isosurface plots Vector plots and streamlines (streamlines are the lines whose tangent direction is the same as the velocity vectors) Animations 36 .

Post-processing (Uncertainty Assessment) • Si l ti error: the difference between a simulation result Simulation S and the truth T (objective reality). hen pe mit * sign and magnitude Delta δ SN of the simulation numerical error itself and the uncertainties in that error estimate UScN 2 2 2 U SN = U I2 + U G + U T2 + U P • Validation: process for assessing simulation modeling uncertainty USM by using benchmark experimental data and. estimating the nce tainties and. estimating the sign and magnitude of the modeling error δSM itself. y y g p when conditions permit. 2 2 2 U V = U D + U SN E = D − S = δ − (δ + δ ) D SM SN E < UV Validation achieved 37 . assumed composed of additive modeling δSM and numerical δSN errors: δ S = S − T = δ SM + δ SN 2 2 2 U S = U SM + U SN • Verification: process for assessing simulation numerical δ SN = δ I + δ G + δ T + δ P = δ I + ∑ δ j j =1 J uncertainties USN and when conditions permit.

Post-processing (UA. Oscillatory divergence: Rk<0. | Rk|>1 • Grid refinement ratio: uniform ratio of grid spacing between meshes. Verification) • C Convergence studies: C t di Convergence studies require a t di i minimum of m=3 solutions to evaluate convergence with respective to input parameters. Consider the solutions ∧ ∧ ∧ corresponding to fine S k 1 . | (iii). meshes rk = Δxk 2 Δxk1 = Δxk3 Δxk2 = Δxk m Δxkm−1 38 .and coarse meshes S k 3 and ε k 21 = Sk 2 − Sk1 ∧ ∧ ε k 32 = Sk 3 − S k 2 ∧ ∧ Rk = ε k 21 ε k 32 (i). Monotonic convergence: 0<Rk<1 ( ) (ii). | Rk|<1 y g . medium S k 2 . Monotonic divergence: Rk>1 (iv). Oscillatory Convergence: Rk<0.

•The number of order magnitude drop and final level of solution residual can be used to determine stopping criteria for iterative solution techniques (1) Oscillatory (2) Convergent (3) Mixed oscillatory/convergent (a) (b) •Typical CFD solution techniques for obtaining steady state solutions UI = 1 ( SU − S L ) 2 Iteration history for series 60: (a). Solution change (b) magnified view of total i hi f i ( ) l i h ifi d i f l resistance over last two periods of oscillation (Oscillatory iterative convergence) 39 .Post-processing (Verification: Iterative Convergence) involve beginning with an initial guess and performing time marching or iteration until a steady state solution is achieved.

and the validity of the assumptions made in RE theory 40 . • The accuracy of the estimates depends on how many terms are retained in the expansion. the magnitude (importance) of the higher-order terms. generalized RE is used to estimate the error δ*k and order of accuracy pk due to the selection of the kth input parameter. • The error is expanded in a power series expansion with integer powers of Δxk as a finite sum.Post-processing (Verification. RE) p g( . ) • Generalized Richardson Extrapolation (RE): For monotonic convergence.

j ≠ k ∑δ * jm (i pk ) order of accuracy for the ith term ˆ S k1 = SC + Δxk1 ( ) ( pk1) g (1) k + + j =1. j ≠ k ∑δ J ˆ S k2 = SC + rk Δxk1 * j3 ( ) ( pk1) g (1) k + j =1. RE) * δ SN = δ SN + ε SN εSN is the error in the estimate * SC is the numerical benchmark S C = S − δ SN Power series expansion J * * ˆ S k = S k − δ I = SC + δ k + ∑ δ * jm m m km m Finite sum for the kth parameter and mth solution d h l i j =1. j ≠ k ∑δ J * j2 ˆ S k3 = SC + rk2 Δxk1 ( ) ( pk1) g (1) k j =1.Post-processing (Verification. j ≠ k J δ k* = ∑ (Δxk n m i =1 m ) ( p ki ) ( g ki ) ˆ Skm = SC + ∑ Δxkm i =1 n ( ) ( pk i ) g + J * j1 (i ) k j =1. j ≠ k ∑δ = Three equations with three unknowns pk = ln ε k32 ε k21 ln(rk ) ( ) δ =δ * k1 * REk 1 εk 21 rkpk − 1 41 .

125 1 − Ck ≥ 0. such as “Uk” will be “Ug” 2.U kc is the uncertainties based on numerical benchmark SC * U kc = (Fs − 1) δ REk 1 • Oscillatory Convergence: Uncertainties can be estimated.25 pkest Ck is the theoretical order of accuracy. all the variables with subscribe symbol k will be replaced by g.6 (1 − C )2 + 1.1]δ REk 1 U kc = ⎨ [ 1−C +1] δ * * δ1 ⎩[|1 −kCk |] |REkRE | k1 1 − Ck < 0.125 * δ RE = k1 rk − 1 pk 1 − Ck < 0. GCI approach 42 . So.Post-processing (UA. only grid uncertainties studied. U k = 1 (SU − S L ) 2 • Divergence • In this course.1⎤ δ * k ⎪ ⎦ REk 1 U k = ⎨⎣ ⎤ * ⎪ ⎡ 2 1 − Ck + 1⎦ δ REk 1 ⎣ ⎩ * ⎧ [2. Correction factors ⎧ ⎡9. 2 for o de a d o order and 1 for 1st order schemes o de sc e es is the correction factor * U k = Fs δ REk 1 2nd | 1 − C k |≥ 0. Verification.25 U k is the uncertainties based on fine mesh solution.4 (1−Ck )2 + 0. but without signs and magnitudes of the errors. cont’d) • Monotonic Convergence: Generalized Richardson Extrapolation ln ( ε k 32 ε k 21 ) pk = ln ( rk ) Ck = ε k 21 rkpk − 1 r pkest k −1 1.

Asymptotic Range) • Asymptotic Range: For sufficiently small Δxk. • To achieve the asymptotic range for practical y p g p geometry and conditions is usually not possible and m>3 is undesirable from a resources point of view est 43 . pk will be close to the theoretical value pk . the solutions are in the asymptotic range such that higher-order terms are negligible and the (i ) assumption that pk(i ) and g k are independent of Δxk is valid. and the correction factor Ck will be close to 1. • When Asymptotic Range reached.Post-processing (Verification.

cont’d) • Verification for velocity profile using AFD: To avoid illRG = ε G21 2 defined ratios. 44 . only grid errors. AFD E can be plot with +Ug and –Ug and +Ugc and –Ugc to see if solution was Ug. NOTE: For verification using AFD for axial velocity profile in laminar pipe flow (CFD Lab1). So. E. the difference between CFD and AFD. Verification. there is no modeling error. respectively. Ugc verified. L2 norm of the εG21 and εG32 are used to define RG and PG ln ε G εG εG 32 2 pG = ( 32 ln (rG ) 2 21 2 ) Where <> and || ||2 are used to denote a profile-averaged quantity (with ratio of solution changes based on L2 norms) and L2 norm respectively norm.Post-processing (UA.

Uv. i l th th lid ti t i t U • Interpretation of the results of a validation effort E < U V Validation achieved U V < E Validation not achieved E = D − S = δ D − (δ SM + δ SN ) 2 2 U V = U SN + U D • Validation example Example: Grid study and validation of wave profile for series 60 45 . the comparison error. E.Post-processing (UA. E is less than the validation uncertainty. Validation) • Validation procedure: simulation modeling uncertainties was presented where for successful validation.

Example of CFD Process using CFD educational interface (Geometry) • Turbulent flows (Re=143K) around Clarky airfoil with angle of attack 6 degree is simulated simulated. • “C” shape domain is applied • The radius of the domain Rc and downstream length Lo should be specified in such a way that the domain size will not affect the simulation results 46 .

Example of CFD Process (Physics) No heat transfer 47 .

Example of CFD Process (Mesh) Grid need to be refined near the foil f f il surface to resolve the boundary l h b d layer 48 .

iteration 49 .Example of CFD Process (Solve) Residuals vs.

Example of CFD Process (Reports) 50 .

Example of CFD Process (Post-processing) 51 .

edu/~me_160 52 .58:160 CFD Labs Schedule CFD Lab Date Lab1: Pipe Flow Sept.2.com/ • Visit class website for more information http://css. 13 Lab4: Ahmed car Nov.fluent.flowlab.10 http://www. 5 Lab 2: Airfoil Flow Sept.engineering. 10 • CFD Labs instructed by Tao Xing and Maysam Mousaviraad • Use the CFD educational interface — FlowLab 1.uiowa. 22 Lab3: Diffuser Oct.