This action might not be possible to undo. Are you sure you want to continue?

# FLUENT TUTORIALS - Cornell University

**About t he FLUENT Tut orials
**

What is FLUENT?

How t o use t hese t ut orials

Syst em requirement s

Convent ions used

Please send us feedback

List of Tutorials

These t ut orials progress from simple t o more complex. I f you are unfamiliar

wit h FLUENT, please begin wit h t he first module.

I nt roduct ion t o CFD Basics

Laminar Pipe Flow

Turbulent Pipe Flow

Compressible Flow in a Nozzle

Flow over an airfoil

Forced Convect ion over a Flat

Plat e

http://instruct1.cit.cornell.edu/courses/fluent/ (1 of 4)11/7/2005 6:08:21 PM

FLUENT TUTORIALS - Cornell University

About the FLUENT Tutorials

This FLUENT short course consist s of a set of t ut orials on using FLUENT t o solve

problems in fluid mechanics. The t ut orials lead t he user t hrough t he st eps

involved in solving a select ed set of problems using GAMBI T ( t he preprocessor)

and FLUENT. We not only provide t he solut ion st eps but also t he rat ionale

behind t hem. I t is wort hwhile for t he user t o underst and t he underlying

concept s as she goes t hrough t he t ut orials in order t o be able t o correct ly apply

FLUENT t o ot her problems. The user would be ill- served by clicking t hrough t he

t ut orials in zombie- mode. Each t ut orial is followed by problems which are

geared t owards st rengt hening and reinforcing t he knowledge and

underst anding gained in t he t ut orials. Working t hrough t he problem set s is an

int rinsic part of t he learning process and shouldn' t be skipped.

These t ut orials have been developed by t he Swanson Engineering Simulat ion

Program in t he Sibley School of Mechanical and Aerospace Engineering at

Cornell Universit y. The Swanson Engineering Simulat ion Program has been

est ablished wit h t he goal of int egrat ing comput er- based simulat ions int o t he

mechanical engineering curriculum. The development of t hese t ut orials is being

support ed by a Facult y I nnovat ion in Teaching award from Cornell Universit y.

What i s FLUENT

FLUENT is a comput at ional fluid dynamics ( CFD) soft ware package t o simulat e

fluid flow problems. I t uses t he finit e- volume met hod t o solve t he governing

equat ions for a fluid. I t provides t he capabilit y t o use different physical models

such as incompressible or compressible, inviscid or viscous, laminar or

t urbulent , et c. Geomet ry and grid generat ion is done using GAMBI T which is t he

preprocessor bundled wit h FLUENT.

How t o use t hese t ut or i al s

These t ut orials are designed t o be used online and run side- by- side wit h t he

FLUENT soft ware. Aft er you launch t he web t ut orials and FLUENT, you will have

t o drag t he browser window t o t he widt h of t he largest image ( about 350

pixels) . To make best use of screen real est at e, move t he windows around and

resize t hem so t hat you approximat e t his screen arrangement .

http://instruct1.cit.cornell.edu/courses/fluent/ (2 of 4)11/7/2005 6:08:21 PM

FLUENT TUTORIALS - Cornell University

Sy st em and sof t w ar e r equi r ement s

q Syst em: Any syst em t hat can run GAMBI T, FLUENT, and a web browser.

q Screen: Resolut ion should be at least 1280 x 1024 pixels for opt imal

viewing. A 17" monit or or larger is recommended.

q GAMBI T version 2. 0. These t ut orials were creat ed using GAMBI T 2. 0.

q FLUENT version 6. 0. These t ut orials were creat ed using FLUENT 6. 0.

q Web Browser: These t ut orials work best in 5. 0 or higher versions of

I nt ernet Explorer and Net scape because st yle sheet support is needed.

These t ut orials can be used wit h Net scape 4. x but may not render

correct ly.

Choose a t ut orial by select ing from t he list at t he t op of t his page

Conv ent i ons used

Each t ut orial begins wit h a problem specificat ion. A solut ion can be obt ained by

following t hese nine st eps:

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Set Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

These st eps appear at t he t op of each page of t he t ut orial wit h t he current st ep

highlight ed in red.

GAMBI T and FLUENT uses cascading menus which are represent ed as follows:

Ma in Men u > File > Exp or t > Mes h . . .

This means t hat in t he Main Menu, click on File. Then, in t he File menu t hat

comes up, click on Export and so on.

http://instruct1.cit.cornell.edu/courses/fluent/ (3 of 4)11/7/2005 6:08:21 PM

FLUENT TUTORIALS - Cornell University

Names of windows are in it alics.

I t ems and opt ions appearing wit hin menus and dialog boxes are pur pl e, i t al i c,

and bol d.

Text and numbers t hat need t o be ent ered are indicat ed in Courier font.

Addit ional explanat ions and relat ed discussions are enclosed in a box.

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering

Feedback .

http://instruct1.cit.cornell.edu/courses/fluent/ (4 of 4)11/7/2005 6:08:21 PM

Fluent Tutorial - Introduction to CFD Basics

I nt r oduct i on t o CFD Basi cs

Aut hor: Raj esh Bhaskaran

E- mail: rb88@cornell. edu

Introduction to CFD Basics

You can download t he following t ut orials in PDF format . You will need Adobe

Acrobat t o read t hese files.

I nt roduct ion t o CFD Basics

Problem set on CFD Basics

Back t o: FLUENT Home Page

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/cfd/index.htm11/7/2005 6:11:18 PM

Introduction to CFD Basics

Rajesh Bhaskaran

Lance Collins

Jan. 2003

This is a quick introduction to the basic concepts underlying CFD. The concepts are

illustrated by applying them to a simple 1D example. We discuss the following topics brieﬂy:

1. The Need for CFD

2. Applications of CFD

3. The Strategy of CFD

4. Discretization Using the Finite-Diﬀerence Method

5. Discretization Using The Finite-Volume Method

6. Assembly of Discrete System and Application of Boundary Conditions

7. Solution of Discrete System

8. Grid Convergence

9. Dealing with Nonlinearity

10. Direct and Iterative Solvers

11. Iterative Convergence

12. Numerical Stability

1

Applications of CFD

CFD is useful in a wide variety of applications and here we note a few to give you an idea of

its use in industry. The simulations shown below have been performed using the FLUENT

software.

CFD can be used to simulate the ﬂow over a vehicle. For instance, it can be used to study

the interaction of propellers or rotors with the aircraft fuselage The following ﬁgure shows

the prediction of the pressure ﬁeld induced by the interaction of the rotor with a helicopter

fuselage in forward ﬂight. Rotors and propellers can be represented with models of varying

complexity.

The temperature distribution obtained from a CFD analysis of a mixing manifold is shown

below. This mixing manifold is part of the passenger cabin ventilation system on the Boeing

767. The CFD analysis showed the eﬀectiveness of a simpler manifold design without the

need for ﬁeld testing.

Bio-medical engineering is a rapidly growing ﬁeld and uses CFD to study the circulatory and

respiratory systems. The following ﬁgure shows pressure contours and a cutaway view that

reveals velocity vectors in a blood pump that assumes the role of heart in open-heart surgery.

CFD is attractive to industry since it is more cost-eﬀective than physical testing. However,

one must note that complex ﬂow simulations are challenging and error-prone and it takes a

lot of engineering expertise to obtain validated solutions.

2

The Strategy of CFD

Broadly, the strategy of CFD is to replace the continuous problem domain with a discrete

domain using a grid. In the continuous domain, each ﬂow variable is deﬁned at every point

in the domain. For instance, the pressure p in the continuous 1D domain shown in the ﬁgure

below would be given as

p = p(x), 0 < x < 1

In the discrete domain, each ﬂow variable is deﬁned only at the grid points. So, in the

discrete domain shown below, the pressure would be deﬁned only at the N grid points.

p

i

= p(x

i

), i = 1, 2, . . . , N

Continuous Domain Discrete Domain

x=0

x=1

x

1

x

i

x

N

0 ≤ x ≤ 1 x = x

1

, x

2

, …,x

N

Grid point

Coupled PDEs + boundary

conditions in continuous

variables

Coupled algebraic eqs. in

discrete variables

In a CFD solution, one would directly solve for the relevant ﬂow variables only at the grid

points. The values at other locations are determined by interpolating the values at the grid

points.

The governing partial diﬀerential equations and boundary conditions are deﬁned in terms

of the continuous variables p,

V etc. One can approximate these in the discrete domain in

terms of the discrete variables p

i

,

V

i

etc. The discrete system is a large set of coupled,

algebraic equations in the discrete variables. Setting up the discrete system and solving it

(which is a matrix inversion problem) involves a very large number of repetitive calculations

and is done by the digital computer.

This idea can be extended to any general problem domain. The following ﬁgure shows

the grid used for solving the ﬂow over an airfoil.

3

Discretization Using the Finite-Diﬀerence Method

To keep the details simple, we will illustrate the fundamental ideas underlying CFD by

applying them to the following simple 1D equation:

du

dx

+ u

m

= 0; 0 ≤ x ≤ 1; u(0) = 1 (1)

We’ll ﬁrst consider the case where m = 1 when the equation is linear. We’ll later consider

the m = 2 case when the equation is nonlinear.

We’ll derive a discrete representation of the above equation with m = 1 on the following

grid:

x

1

=0 x

2

=1/3 x

3

=2/3 x

4

=1

∆x=1/3

This grid has four equally-spaced grid points with ∆x being the spacing between successive

points. Since the governing equation is valid at any grid point, we have

du

dx

i

+ u

i

= 0 (2)

where the subscript i represents the value at grid point x

i

. In order to get an expression for

(du/dx)

i

in terms of u at the grid points, we expand u

i−1

in a Taylor’s series:

u

i−1

= u

i

−∆x

du

dx

i

+ O(∆x

2

)

Rearranging gives

du

dx

i

=

u

i

−u

i−1

∆x

+ O(∆x) (3)

The error in (du/dx)

i

due to the neglected terms in the Taylor’s series is called the truncation

error. Since the truncation error above is O(∆x), this discrete representation is termed ﬁrst-

order accurate.

Since the error in (du/dx)

i

due to the neglected terms in the Taylor’s series is of O(∆x),

this representation is termed as ﬁrst-order accurate. Using (3) in (2) and excluding higher-

order terms in the Taylor’s series, we get the following discrete equation:

u

i

−u

i−1

∆x

+ u

i

= 0 (4)

Note that we have gone from a diﬀerential equation to an algebraic equation!

This method of deriving the discrete equation using Taylor’s series expansions is called

the ﬁnite-diﬀerence method. However, most commercial CFD codes use the ﬁnite-volume or

ﬁnite-element methods which are better suited for modeling ﬂow past complex geometries.

For example, the FLUENT code uses the ﬁnite-volume method whereas ANSYS uses the

ﬁnite-element method. We’ll brieﬂy indicate the philosophy of the ﬁnite-volume method

next but will keep using the ﬁnite-diﬀerence approach to illustrate the underlying concepts

since they are very similar between the diﬀerent approaches with the ﬁnite-diﬀerence method

being easier to understand.

4

Discretization Using The Finite-Volume Method

If you look closely at the airfoil grid shown earlier, you’ll see that it consists of quadrilaterals.

In the ﬁnite-volume method, such a quadrilateral is commonly referred to as a “cell” and a

grid point as a “node”. In 2D, one could also have triangular cells. In 3D, cells are usually

hexahedrals, tetrahedrals, or prisms. In the ﬁnite-volume approach, the integral form of the

conservation equations are applied to the control volume deﬁned by a cell to get the discrete

equations for the cell. For example, the integral form of the continuity equation was given

earlier. For steady, incompressible ﬂow, this equation reduces to

S

V· ˆ ndS = 0 (5)

The integration is over the surface S of the control volume and ˆ n is the outward normal

at the surface. Physically, this equation means that the net volume ﬂow into the control

volume is zero.

Consider the rectangular cell shown below.

face 1

(u

1

,v

1

)

face 2

face 3

face 4

(u

2

,v

2

)

(u

3

,v

3

)

(u

4

,v

4

)

Cell center

∆x

x

y

∆y

The velocity at face i is taken to be

V

i

= u

i

ˆ

i + v

i

ˆ

j. Applying the mass conservation

equation (5) to the control volume deﬁned by the cell gives

−u

1

∆y −v

2

∆x + u

3

∆y + v

4

∆x = 0

This is the discrete form of the continuity equation for the cell. It is equivalent to summing

up the net mass ﬂow into the control volume and setting it to zero. So it ensures that the net

mass ﬂow into the cell is zero i.e. that mass is conserved for the cell. Usually the values at

the cell centers are stored. The face values u

1

, v

2

, etc. are obtained by suitably interpolating

the cell-center values for adjacent cells.

Similarly, one can obtain discrete equations for the conservation of momentum and energy

for the cell. One can readily extend these ideas to any general cell shape in 2D or 3D and any

conservation equation. Take a few minutes to contrast the discretization in the ﬁnite-volume

approach to that in the ﬁnite-diﬀerence method discussed earlier.

Look back at the airfoil grid. When you are using FLUENT or another ﬁnite-volume code,

it’s useful to remind yourself that the code is ﬁnding a solution such that mass, momentum,

energy and other relevant quantities are being conserved for each cell.

5

Assembly of Discrete System and Application of Boundary Condi-

tions

Recall that the discrete equation that we obtained using the ﬁnite-diﬀerence method was

u

i

−u

i−1

∆x

+ u

i

= 0

Rearranging, we get

−u

i−1

+ (1 + ∆x)u

i

= 0

Applying this equation to the 1D grid shown earlier at grid points i = 2, 3, 4 gives

−u

1

+ (1 + ∆x) u

2

= 0 (i = 2) (6)

−u

2

+ (1 + ∆x) u

3

= 0 (i = 3) (7)

−u

3

+ (1 + ∆x) u

4

= 0 (i = 4) (8)

The discrete equation cannot be applied at the left boundary (i=1) since u

i−1

is not deﬁned

here. Instead, we use the boundary condition to get

u

1

= 1 (9)

Equations (6)-(9) form a system of four simultaneous algebraic equations in the four

unknowns u

1

, u

2

, u

3

and u

4

. It’s convenient to write this system in matrix form:

1 0 0 0

−1 1 + ∆x 0 0

0 −1 1 + ∆x 0

0 0 −1 1 + ∆x

¸

¸

¸

¸

¸

u

1

u

2

u

3

u

4

¸

¸

¸

¸

¸

=

1

0

0

0

¸

¸

¸

¸

¸

(10)

In a general situation, one would apply the discrete equations to the grid points (or cells

in the ﬁnite-volume method) in the interior of the domain. For grid points (or cells) at or

near the boundary, one would apply a combination of the discrete equations and boundary

conditions. In the end, one would obtain a system of simultaneous algebraic equations with

the number of equations being equal to the number of independent discrete variables. The

process is essentially the same as above with the details being much more complex.

FLUENT, like other commercial CFD codes, oﬀers a variety of boundary condition op-

tions such as velocity inlet, pressure inlet, pressure outlet, etc. It is very important that

you specify the proper boundary conditions in order to have a well-deﬁned problem. Also,

read through the documentation for a boundary condition option to understand what it

does before you use it (it might not be doing what you expect). A single wrong boundary

condition can give you a totally wrong result.

6

Solution of Discrete System

The discrete system (10) for our own humble 1D example can be easily inverted to obtain

the unknowns at the grid points. Solving for u

1

, u

2

, u

3

and u

4

in turn and using ∆x = 1/3,

we get

u

1

= 1 u

2

= 3/4 u

3

= 9/16 u

4

= 27/64

The exact solution for the 1D example is easily calculated to be

u

exact

= exp(−x)

The ﬁgure below shows the comparison of the discrete solution obtained on the four-point

grid with the exact solution. The error is largest at the right boundary where it is equal to

14.7%.

0 0.2 0.4 0.6 0.8 1

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

x

u

Numerical solution

Exact solution

In a practical CFD application, one would have thousands to millions of unknowns in the

discrete system and if one uses, say, a Gaussian elimination procedure naively to invert the

matrix, it would be take the computer forever to perform the calculation. So a lot of work

goes into optimizing the matrix inversion in order to minimize the CPU time and memory

required. The matrix to be inverted is sparse i.e. most of the entries in it are zeros since the

discrete equation at a grid point or cell will contain only quantities from the neighboring

points or cells. A CFD code would store only the non-zero values to minimize memory

usage. It would also generally use an iterative procedure to invert the matrix; the longer one

iterates, the closer one gets to the true solution for the matrix inversion.

7

Grid Convergence

While developing the ﬁnite-diﬀerence approximation for the 1D example, we saw that the

truncation error in our discrete system is O(∆x). So one expects that as the number of grid

points is increased and ∆x is reduced, the error in the numerical solution would decrease

and the agreement between the numerical and exact solutions would get better.

Let’s consider the eﬀect of increasing the number of grid points N on the numerical

solution of the 1D problem. We’ll consider N = 8 and N = 16 in addition to the N = 4

case solved previously. We can easily repeat the assembly and solution steps for the discrete

system on each of these additional grids. The following ﬁgure compares the results obtained

on the three grids with the exact solution. As expected, the numerical error decreases as the

number of grid points is increased.

0 0.2 0.4 0.6 0.8 1

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

x

u

N=4

N=8

N=16

Exact solution

When the numerical solutions obtained on diﬀerent grids agree to within a level of tolerance

speciﬁed by the user, they are referred to as “grid converged” solutions. The concept of

grid convergence applies to the ﬁnite-volume approach also where the numerical solution, if

correct, becomes independent of the grid as the cell size is reduced. It is very important

that you investigate the eﬀect of grid resolution on the solution in every CFD problem you

solve. Never trust a CFD solution unless you have convinced yourself that the solution is

grid converged to an acceptance level of tolerance (which would be problem dependent).

8

Dealing with Nonlinearity

The momentum conservation equation for a ﬂuid is nonlinear due to the convection term

(

V · ∇)

**V . Phenomena such as turbulence and chemical reaction introduce additional non-
**

linearities. The highly nonlinear nature of the governing equations for a ﬂuid makes it

challenging to obtain accurate numerical solutions for complex ﬂows of practical interest.

We will demonstrate the eﬀect of nonlinearity by setting m = 2 in our simple 1D exam-

ple (1):

du

dx

+ u

2

= 0; 0 ≤ x ≤ 1; u(0) = 1

A ﬁrst-order ﬁnite-diﬀerence approximation to this equation, analogous to that in (4) for

m = 1, is

u

i

−u

i−1

∆x

+ u

2

i

= 0 (11)

This is a nonlinear algebraic equation with the u

2

i

term being the source of the nonlinearity.

The strategy that is adopted to deal with nonlinearity is to linearize the equations about

a guess value of the solution and to iterate until the guess agrees with the solution to a

speciﬁed tolerance level. We’ll illustrate this on the above example. Let u

g

i

be the guess for

u

i

. Deﬁne

∆u

i

= u

i

−u

g

i

Rearranging and squaring this equation gives

u

2

i

= u

2

g

i

+ 2u

g

i

∆u

i

+ (∆u

i

)

2

Assuming that ∆u

i

u

g

i

, we can neglect the ∆u

2

i

term to get

u

2

i

u

2

g

i

+ 2u

g

i

∆u

i

= u

2

g

i

+ 2u

g

i

(u

i

−u

g

i

)

Thus,

u

2

i

2u

g

i

u

i

−u

2

g

i

The ﬁnite-diﬀerence approximation (11) after linearization becomes

u

i

−u

i−1

∆x

+ 2u

g

i

u

i

−u

2

g

i

= 0 (12)

Since the error due to linearization is O(∆u

2

), it tends to zero as u

g

→u.

In order to calculate the ﬁnite-diﬀerence approximation (12), we need guess values u

g

at

the grid points. We start with an initial guess value in the ﬁrst iteration. For each subsequent

iteration, the u value obtained in the previous iteration is used as the guess value.

Iteration 1: u

(1)

g

= Initial guess

Iteration 2: u

(2)

g

= u

(1)

.

.

.

Iteration l: u

(l)

g

= u

(l−1)

The superscript indicates the iteration level. We continue the iterations until they converge.

We’ll defer the discussion on how to evaluate convergence until a little later.

This is essentially the process used in CFD codes to linearize the nonlinear terms in the

conservations equations, with the details varying depending on the code. The important

points to remember are that the linearization is performed about a guess and that it is

necessary to iterate through successive approximations until the iterations converge.

9

Direct and Iterative Solvers

We saw that we need to perform iterations to deal with the nonlinear terms in the governing

equations. We next discuss another factor that makes it necessary to carry out iterations in

practical CFD problems.

Verify that the discrete equation system resulting from the ﬁnite-diﬀerence approxima-

tion (12) on our four-point grid is

1 0 0 0

−1 1 + 2∆xu

g

2

0 0

0 −1 1 + 2∆xu

g

3

0

0 0 −1 1 + 2∆xu

g

4

¸

¸

¸

¸

¸

u

1

u

2

u

3

u

4

¸

¸

¸

¸

¸

=

1

∆xu

2

g

2

∆xu

2

g

3

∆xu

2

g

4

¸

¸

¸

¸

¸

(13)

In a practical problem, one would usually have millions of grid points or cells so that each

dimension of the above matrix would be of the order of a million (with most of the elements

being zeros). Inverting such a matrix directly would take a prohibitively large amount of

memory. So instead, the matrix is inverted using an iterative scheme as discussed below.

Rearrange the ﬁnite-diﬀerence approximation (12) at grid point i so that u

i

is expressed

in terms of the values at the neighboring grid points and the guess values:

u

i

=

u

i−1

+ ∆xu

2

g

i

1 + 2 ∆xu

g

i

If a neighboring value at the current iteration level is not available, we use the guess value

for it. Let’s say that we sweep from right to left on our grid i.e. we update u

4

, then u

3

and

ﬁnally u

2

in each iteration. In the m

th

iteration, u

(l)

i−1

is not available while updating u

m

i

and

so we use the guess value u

(l)

g

i−1

for it instead:

u

(l)

i

=

u

(l)

g

i−1

+ ∆xu

(l)

2

g

i

1 + 2 ∆xu

(l)

g

i

(14)

Since we are using the guess values at neighboring points, we are eﬀectively obtaining only

an approximate solution for the matrix inversion in (13) during each iteration but in the

process have greatly reduced the memory required for the inversion. This tradeoﬀ is good

strategy since it doesn’t make sense to expend a great deal of resources to do an exact matrix

inversion when the matrix elements depend on guess values which are continuously being

reﬁned. In an act of cleverness, we have combined the iteration to handle nonlinear terms

with the iteration for matrix inversion into a single iteration process. Most importantly, as

the iterations converge and u

g

→u, the approximate solution for the matrix inversion tends

towards the exact solution for the inversion since the error introduced by using u

g

instead

of u in (14) also tends to zero.

Thus, iteration serves two purposes:

1. It allows for eﬃcient matrix inversion with greatly reduced memory requirements.

2. It is necessary to solve nonlinear equations.

In steady problems, a common and eﬀective strategy used in CFD codes is to solve the

unsteady form of the governing equations and “march” the solution in time until the solution

converges to a steady value. In this case, each time step is eﬀectively an iteration, with the

the guess value at any time level being given by the solution at the previous time level.

10

Iterative Convergence

Recall that as u

g

→ u, the linearization and matrix inversion errors tends to zero. So we

continue the iteration process until some selected measure of the diﬀerence between u

g

and

u, refered to as the residual, is “small enough”. We could, for instance, deﬁne the residual

R as the RMS value of the diﬀerence between u and u

g

on the grid:

R ≡

N

¸

i=1

(u

i

−u

g

i

)

2

N

It’s useful to scale this residual with the average value of u in the domain. An unscaled

residual of, say, 0.01 would be relatively small if the average value of u in the domain is 5000

but would be relatively large if the average value is 0.1. Scaling ensures that the residual is

a relative rather than an absolute measure. Scaling the above residual by dividing by the

average value of u gives

R =

¸

¸

¸

¸

¸

¸

¸

N

¸

i=1

(u

i

−u

g

i

)

2

N

¸

¸

¸

¸

¸

¸

¸

N

N

¸

i=1

u

i

¸

=

N

N

¸

i=1

(u

i

−u

g

i

)

2

N

¸

i=1

u

i

(15)

For the nonlinear 1D example, we’ll take the initial guess at all grid points to be equal

to the value at the left boundary i.e. u

(1)

g

= 1. In each iteration, we update u

g

, sweep

from right to left on the grid updating, in turn, u

4

, u

3

and u

2

using (14) and calculate

the residual using (15). We’ll terminate the iterations when the residual falls below 10

−9

(which is referred to as the convergence criterion). Take a few minutes to implement this

procedure in MATLAB which will help you gain some familiarity with the mechanics of the

implementation. The variation of the residual with iterations obtained from MATLAB is

shown below. Note that logarithmic scale is used for the ordinate. The iterative process

converges to a level smaller than 10

−9

in just 6 iterations. In more complex problems, a lot

more iterations would be necessary for achieving convergence.

1 2 3 4 5 6

10

−10

10

−8

10

−6

10

−4

10

−2

10

0

Iteration number

R

e

s

i

d

u

a

l

11

The solution after 2,4 and 6 iterations and the exact solution are shown below in the

right ﬁgure. It can easily be veriﬁed that the exact solution is given by

u

exact

=

1

x + 1

The solutions for iterations 4 and 6 are indistinguishable on the graph. This is another

indication that the solution has converged. The converged solution doesn’t agree well with

the exact solution because we are using a coarse grid for which the truncation error is

relatively large. The iterative convergence error, which is of order 10

−9

, is swamped out

by the truncation error of order 10

−1

. So driving the residual down to 10

−9

when the

truncation error is of order 10

−1

is a waste of computing resources. In a good calculation,

both errors would be of comparable level and less than a tolerance level chosen by the user.

The agreement between the numerical and exact solutions should get much better on reﬁning

the grid as was the case for m = 1.

0 0.2 0.4 0.6 0.8 1

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

x

u

Iteration 2

Iteration 4

Iteration 6

Exact

Some points to note:

1. Diﬀerent codes use slightly diﬀerent deﬁnitions for the residual. Read the documenta-

tion to understand how the residual is calculated.

2. In the FLUENT code, residuals are reported for each conservation equation. A discrete

conservation equation at any cell can be written in the form LHS = 0. For any iteration,

if one uses the current solution to compute the LHS, it won’t be exactly equal to

zero, with the deviation from zero being a mesaure of how far one is from achieving

convergence. So FLUENT calculates the residual as the (scaled) mean of the absolute

value of the LHS over all cells.

3. The convergence criterion you choose for each conservation equation is problem- and

code-dependent. It’s a good idea to start with the default values in the code. One may

then have to tweak these values.

12

Numerical Stability

In our previous 1D example, the iterations converged very rapidly with the residual falling

below the convergence criterion of 10

−9

in just 6 iterations. In more complex problems, the

iterations converge more slowly and in some instances, may even diverge. One would like

to know a priori the conditions under which a given numerical scheme converges. This is

determined by performing a stability analysis of the numerical scheme. A numerical method

is referred to as being stable when the iterative process converges and as being unstable

when it diverges. It is not possible to carry out an exact stability analysis for the Euler or

Navier-Stokes equations. But a stability analysis of simpler, model equations provides useful

insight and approximate conditions for stability. As mentioned earlier, a common strategy

used in CFD codes for steady problems is to solve the unsteady equations and march in time

until the solution converges to a steady state. A stability analysis is usually performed in

the context of time-marching.

While using time-marching to a steady state, we are only interested in accurately obtain-

ing the asymptotic behavior at large times. So we would like to take as large a time-step

∆t as possible to reach the steady state in the least number of time-steps. There is usually

a maximum allowable time-step ∆t

max

beyond which the numerical scheme is unstable. If

∆t > ∆t

max

, the numerical errors will grow exponentially in time causing the solution to

diverge from the steady-state result. The value of ∆t

max

depends on the numerical dis-

cretization scheme used. There are two classes of numerical shemes, explicit and implicit,

with very diﬀerent stability characteristics which we’ll brieﬂy discuss next.

Explicit and Implicit Schemes

The diﬀerence between explicit and implicit schemes can be most easily illustrated by ap-

plying them to the wave equation

∂u

∂t

+ c

∂u

∂x

= 0

where c is the wavespeed. One possible way to discretize this equation at grid point i and

time-level n is

u

n

i

−u

n−1

i

∆t

+ c

u

n−1

i

−u

n−1

i−1

∆x

= O(∆t, ∆x) (16)

The crucial thing to note here is that the spatial derivative is evaluated at the n−1 time-level.

Solving for u

n

i

gives

u

n

i

=

¸

1 −

c∆t

∆x

u

n−1

i

+

c∆t

∆x

u

n−1

i−1

(17)

This is an explicit expression i.e. the value of u

n

i

at any grid point can be calculated directly

from this expression without the need for any matrix inversion. The scheme in (16) is known

as an explicit scheme. Since u

n

i

at each grid point can be updated independently, these

schemes are easy to implement on the computer. On the downside, it turns out that this

scheme is stable only when

C ≡

c∆t

∆x

≤ 1

where C is called the Courant number. This condition is refered to as the Courant-Friedrichs-

Lewy or CFL condition. While a detailed derivation of the CFL condition through stability

analysis is outside the scope of the current discussion, it can seen that the coeﬃcient of u

n−1

i

13

in (17) changes sign depending on whether C > 1 or C < 1 leading to very diﬀerent behavior

in the two cases. The CFL condition places a rather severe limitation on ∆t

max

.

In an implicit scheme, the spatial derivative term is evaluated at the n time-level:

u

n

i

−u

n−1

i

∆t

+ c

u

n

i

−u

n

i−1

∆x

= O(∆t, ∆x)

In this case, we can’t update u

n

i

at each grid point independently. We instead need to solve a

system of algebraic equations in order to calculate the values at all grid points simultaneously.

It can be shown that this scheme is unconditionally stable so that the numerical errors will

be damped out irrespective of how large the time-step is.

The stability limits discussed above apply speciﬁcally to the wave equation. In general,

explicit schemes applied to the Euler or Navier-Stokes equations have the same restriction

that the Courant number needs to be less than or equal to one. Implicit schemes are not

unconditonally stable for the Euler or Navier-Stokes equations since the nonlinearities in

the governing equations often limit stability. However, they allow a much larger Courant

number than explicit schemes. The speciﬁc value of the maximum allowable Courant number

is problem dependent.

Some points to note:

1. CFD codes will allow you to set the Courant number (which is also referred to as

the CFL number) when using time-stepping. Taking larger time-steps leads to faster

convergence to the steady state, so it is advantageous to set the Courant number as

large as possible within the limits of stability.

2. You may ﬁnd that a lower Courant number is required during startup when changes

in the solution are highly nonlinear but it can be increased as the solution progresses.

3. Under-relaxation for non-timestepping

14

Turbulence Modeling

There are two radically diﬀerent states of ﬂows that are easily identiﬁed and distinguished:

laminar ﬂow and turbulent ﬂow. Laminar ﬂows are characterized by smoothly varying ve-

locity ﬁelds in space and time in which individual “laminae” (sheets) move past one another

without generating cross currents. These ﬂows arise when the ﬂuid viscosity is suﬃciently

large to damp out any perturbations to the ﬂow that may occur due to boundary imper-

fections or other irregularities. These ﬂows occur when at low-to-moderate values of the

Reynolds number. In contrast, turbulent ﬂows are characterized by large, nearly random

ﬂuctuations in velocity and pressure in both space and time. These ﬂuctuations arise from

instabilities that grow until nonlinear interactions cause them to break down into ﬁner and

ﬁner whirls that eventually are dissipated (into heat) by the action of viscosity. Turbulent

ﬂows occur in the opposite limit of high Reynolds numbers.

2.3

2.2

2.1

2.0

1.9

1.8

1.7

y

l

a

b

e

l

100 80 60 40 20 0

xlabel

(a)

PSfrag replacements

u

t

-0.4

-0.2

0.0

0.2

y

l

a

b

e

l

100 80 60 40 20 0

xlabel

(b)

PSfrag replacements

u

t

u

t

0.12

0.10

0.08

0.06

0.04

0.02

0.00

y

l

a

b

e

l

100 80 60 40 20 0

xlabel

(c)

PSfrag replacements

u

2

t

Figure 1: Example of a time history of a component of a ﬂuctuating velocity at a point in

a turbulent ﬂow. (a) Shows the velocity, (b) shows the ﬂuctuating component of velocity

u

**≡ u −u and (c) shows the square of the ﬂuctuating velocity. Dashed lines in (a) and (c)
**

indicate the time averages.

A typical time history of the ﬂow variable u at a ﬁxed point in space is shown in Fig. 1(a).

The dashed line through the curve indicates the “average” velocity. We can deﬁne three types

of averages:

1. Time average

15

2. Volume average

3. Ensemble average

The most mathematically general average is the ensemble average, in which you repeat a

given experiment a large number of times and average the quantity of interest (say velocity)

at the same position and time in each experiment. For practical reasons, this is rarely done.

Instead, a time or volume average (or combination of the two) is made with the assumption

that they are equivalent to the ensemble average. For the sake of this discussion, let us deﬁne

the time average for a stationary ﬂow

1

as

u(y) ≡ lim

τ→∞

1

2τ

τ

−τ

u(y, t)dt (18)

The deviation of the velocity from the mean value is called the ﬂuctuation and is usually

deﬁned as

u

≡ u −u (19)

Note that by deﬁnition u

**= 0 (the average of the ﬂuctuation is zero). Consequently, a
**

better measure of the strength of the ﬂuctuation is the average of the square of a ﬂuctuating

variable. Figures 1(b) and 1(c) show the time evolution of the velocity ﬂuctuation, u

, and

the square of that quantity, u

2

. Notice that the latter quantity is always greater than zero

as is its average.

The equations governing a turbulent ﬂow are precisely the same as for a laminar ﬂow;

however, the solution is clearly much more complicated in this regime. The approaches to

solving the ﬂow equations for a turbulent ﬂow ﬁeld can be roughly divided into two classes.

Direct numerical simulations (DNS) use the speed of modern computers to numerically

integrate the Navier Stokes equations, resolving all of the spatial and temporal ﬂuctuations,

without resorting to modeling. In essence, the solution procedure is the same as for laminar

ﬂow, except the numerics must contend with resolving all of the ﬂuctuations in the velocity

and pressure. DNS remains limited to very simple geometries (e.g., channel ﬂows, jets and

boundary layers) and is extremely expensive to run.

2

The alternative to DNS found in

most CFD packages (including FLUENT) is to solve the Reynolds Averaged Navier Stokes

(RANS) equations. RANS equations govern the mean velocity and pressure. Because these

quantities vary smoothly in space and time, they are much easier to solve; however, as will

be shown below, they require modeling to “close” the equations and these models introduce

signiﬁcant error into the calculation.

To demonstrate the closure problem, we consider fully developed turbulent ﬂow in a

channel of height 2H. Recall that with RANS we are interested in solving for the mean

velocity u(y) only. If we formally average the Navier Stokes equations and simplify for this

geometry we arrive at the following

du

v

dy

+

1

ρ

dp

dx

= ν

d

2

u(y)

dy

2

(20)

1

A stationary ﬂow is deﬁned as one whose statistics are not changing in time. An example of a stationary

ﬂow is steady ﬂow in a channel or pipe.

2

The largest DNS to date was recently published by Kaneda et al., Phys. Fluids 15(2):L21–L24 (2003);

they used 4096

3

grid point, which corresponds roughly to 0.5 terabytes of memory per variable!

16

subject to the boundary conditions

y = 0

du

dy

= 0 , (21)

y = H u = 0 , (22)

The quantity u

v

**, known as the Reynolds stress,
**

3

is a higher-order moment that must

be modeled in terms of the knowns (i.e., u(y) and its derivatives). This is referred to as

the “closure” approximation. The quality of the modeling of this term will determine the

reliability of the computations.

4

Turbulence modeling is a rather broad discipline and an in-depth discussion is beyond

the scope of this introduction. Here we simply note that the Reynolds stress is modeled in

terms of two turbulence parameters, the turbulent kinetic energy k and the turbulent energy

dissipation rate deﬁned below

k ≡

1

2

u

2

+ v

2

+ w

2

(23)

≡ ν

∂u

∂x

2

+

∂u

∂y

2

+

∂u

∂z

2

+

∂v

∂x

2

+

∂v

∂y

2

+

∂v

∂z

2

+

∂w

∂x

2

+

∂w

∂y

2

+

∂w

∂z

2

¸

¸

(24)

where (u

, v

, w

**) is the ﬂuctuating velocity vector. The kinetic energy is zero for laminar
**

ﬂow and can be as large as 5% of the kinetic energy of the mean ﬂow in a highly turbulent

case. The family of models is generally known as k– and they form the basis of most CFD

packages (including FLUENT). We will revisit turbulence modeling towards the end of the

semester.

3

Name after the same Osborne Reynolds from which we get the Reynolds number.

4

Notice that if we neglect the Reynolds stress the equations reduce to the equations for laminar ﬂow;

thus, the Reynolds stress is solely responsible for the diﬀerence in the mean proﬁle for laminar (parabolic)

and turbulent (blunted) ﬂows.

17

Problem Set for “Intro to CFD” Notes

Consider the following diﬀerential equation

d

2

u

dx

2

− 2 u

3

= 0; 0 ≤ x ≤ 9; u(0) = 1, u(9) = 0.1

• Apply the ﬁnite-diﬀerence method to this equation to get a linearized diﬀerence equa-

tion at grid point i away from the boundary. Note that a second-order diﬀerence

approximation for the second-derivative is

d

2

u

dx

2

i

=

u

i−1

− 2u

i

+ u

i+1

∆x

2

+O

∆x

2

**• Assemble the discrete system of equations for a four-point grid into a matrix system
**

of the form

[A]{u} = {b}

where

{u} = {u

1

u

2

u

3

u

4

}

T

• Develop a MATLAB program to solve the ﬁnite-diﬀerence equations on a grid with N

points. Apply this code to obtain the solution on a 4-point grid (∆x = 3). For the

initial guess, use a linear variation between the two boundary values. Converge your

solution until the residual is below 10

−6

. Plot the residuals vs. iteration number.

Hint: In MATLAB, initialize all elements of [A] to zero. For row i of [A] when

2 ≤ i ≤ N − 1, you need to set only the elements A

i,i−1

, A

i,i

and A

i,i+1

.

• Plot the ﬁnite-diﬀerence solution obtained on the 4-point grid and compare it with the

exact solution

u

exact

=

1

x + 1

• Use your MATLAB program to obtain the solution on a 7-point grid (∆x = 1.5).

Plot the solution and compare it with the solution for the 4-point grid and the exact

solution.

1

Fluent Tutorial - Laminar Pipe Flow

Lami nar Pi pe Fl ow

Aut hor: Raj esh Bhaskaran

E- mail: rb88@cornell. edu

Pr obl em Speci f i cat i on

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Problem Specification

Consider fluid flowing t hrough a circular pipe of cont ant cross- sect ion. The pipe

diamet er D= 0. 2 m and lengt h L= 8 m. The inlet velocit y V

in

= 1 m/ s. Consider t he

velocit y t o be const ant over t he inlet cross- sect ion. The fluid exhaust s int o t he

ambient at mosphere which is at a pressure of 1 at m. Take densit y ρ= 1 kg/ m

3

and coefficient of viscosit y µ= 2 x 10

- 3

kg/ ( ms) . The Reynolds number Re based

on t he pipe diamet er is

where Vavg is t he average velocit y at t he inlet , which is 1m/ s in t his case.

Solve t his problem using FLUENT. Plot t he cent erline velocit y, wall skin- frict ion

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/index.htm (1 of 2)11/7/2005 6:32:02 PM

Fluent Tutorial - Laminar Pipe Flow

coefficient , and velocit y profile at t he out let . Validat e your result s.

Not e: The values used for t he inlet velocit y and flow propert ies are chosen for

convenience rat her t han t o reflect realit y. The key paramet er value t o focus on is

t he Reynolds no.

Preliminary Analysis

We expect t he viscous boundary layer t o grow along t he pipe st art ing at t he

inlet . I t will event ually grow t o fill t he pipe complet ely ( provided t hat t he pipe is

long enough) . When t his happens, t he flow becomes fully- developed and t here is

no variat ion of t he velocit y profile in t he axial direct ion, x ( see figure below) . One

can obt ain a closed- form solut ion t o t he governing equat ions in t he fully-

developed region. You should have seen t his in t he I nt roduct ion t o Fluid

Mechanics course. We will compare t he numerical result s in t he fully- developed

region wit h t he corresponding analyt ical result s. So it ' s a good idea for you t o go

back t o your t ext book in t he I nt ro course and review t he fully- developed flow

analysis. What are t he values of cent erline velocit y and frict ion fact or you expect

in t he fully- developed region based on t he analyt ical solut ion? What is t he

solut ion for t he velocit y profile?

We' ll creat e t he geomet ry and mesh in GAMBI T which is t he preprocessor for

FLUENT, and t hen read t he mesh int o FLUENT and solve for t he flow solut ion.

Go t o St ep 1: Creat e Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/index.htm (2 of 2)11/7/2005 6:32:02 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Cr eat e Geomet r y i n GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 1: Create Geometry in GAMBIT

I f you would prefer t o skip t he mesh generat ion st eps, you can creat e a

working direct ory ( see below) , download t he mesh from here ( right click and

save as pipe. msh) int o t he working direct ory and go st raight t o st ep 4.

St r at egy f or Cr eat i ng Geomet r y

I n order t o creat e t he rect angle, we will first creat e t he vert ices at t he four

corners. We' ll t hen j oin adj acent vert ices by st raight lines t o form t he "edges" of

t he rect angle. Last ly, we' ll creat e a "face" corresponding t o t he area enclosed by

t he edges. I n St ep 2, we' ll mesh t he face i. e. t he rect angle. Not e t hat in 3D

problems, you' ll have t o form a "volume" from faces. So t he hierarchy of

geomet ric obj ect s in GAMBI T is vert ices - > edges - > faces - > volumes.

Cr eat e a Wor k i ng Di r ect or y

Creat e a folder called pipe in a convenient locat ion. We' ll use t his as t he working

folder in which files creat ed during t he session will be st ored.

Not e f or ACCEL comput er l ab user s: Each user get s his/ her own 100 MB of

disk space under S: at ACCEL. You can put your files in S: and it ' ll be accessible

from any comput er. This is where you should put files t hat you want t o keep and

access lat er on.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (1 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

St ar t GAMBI T

St art your command prompt .

St a r t > Ru n

I n Windows NT/ 2000/ XP: Type cmd and press ent er.

I n Windows 95/ 98/ ME: Type command and press ent er.

Navigat e your way t o your working folder. For example, if you creat ed a folder

named fluent on drive S: in Windows, t ype cd S:\fluent at t he command

prompt and press Ent er .

St art GAMBI T by t yping gambit -id pipe at t he command prompt and pressing

Ent er .

I f t his doesn' t work, t ry t yping t he full pat hname t o t he GAMBI T execut able:

c:\fluent.inc\ntbin\ntx86\gambit -id pipe

This brings up t he GAMBI T int erface and t ells GAMBI T t o use pipe as t he default

prefix for all files creat ed during t he session. I n Windows, t he Exceed X- server

st art s up before t he GAMBI T int erface comes up. Exceed is a t hird- part y

applicat ion needed t o render t he int erface in Windows ( GAMBI T was originally

developed under Unix) . To make best use of screen real est at e, move t he

windows and resize t hem so t hat you approximat e t his screen arrangement . This

way you can read inst ruct ions in t he browser window and implement t hem in

GAMBI T.

You can resize t he t ext in t he browser window t o your t ast e and comfort :

I n I nt ernet Explorer: Men u ba r > View > Text Size, t hen choose t he appropriat e

font size.

I n Net scape: Men u ba r > View > I n cr ea s e Fon t or Men u ba r > View > Decr ea s e

Fon t .

The GAMBI T I nt erface consist s of t he following:

q Main Menu Bar:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (2 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

Not e t hat t he j ob name pipe appears aft er I D: in t he t it le bar of t he Ut ilit y

Menu.

q Operat ion Toolpad:

We' ll more or less work our way across t he Operat ion Toolpad as we go

t hrough t he solut ion st eps. Not ice t hat as each of t he t op but t ons is

select ed, a different "sub- pad" appears. The Geomet ry sub- pad is shown in

t he above snaphot .

q Global Cont rol Toolpad:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (3 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

The Global Cont rol Toolpad has opt ions such as Fi t t o Scr een and

Undo t hat are very handy during t he course of geomet ry and mesh

creat ion.

q GAMBI T Graphics:

This is t he window where t he graphical result s of operat ions are displayed.

q GAMBI T Descript ion Panel:

The Descript ion Panel cont ains descript ions of but t ons or obj ect s t hat t he

mouse is point ing t o. Move your mouse over some but t ons and not ice t he

corresponding t ext in t he Descript ion Panel.

q GAMBI T Transcript Window:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (4 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

This is t he window t o which out put from GAMBI T commands is writ t en and

which provides feedback on t he act ions t aken by GAMBI T as you perform

operat ions. I f, at some point , you are not sure you clicked t he right but t on

or ent ered a value correct ly, t his is where t o look t o figure out what you

j ust did. You can click on t he arrow but t on in t he upper right hand corner t o

make t he Transcript window full- sized. You can click on t he arrow again t o

ret urn t he window t o it s original size. Go ahead, give t his a t ry.

Sel ect Sol ver

Specify t hat t he mesh t o be creat ed is for use wit h FLUENT 6. 0:

Ma in Men u > Solver > FLUENT 5/ 6

Verify t his has been done by looking in t he Transcript Window where you should

see:

The boundary t ypes t hat you' ll be able t o select in t he t hird st ep depends on

t he solver select ed.

We can assume t hat t he flow is axisymmet ric. The problem domain is:

where r and x are t he radial and axial coordinat es, respect ively.

St r at egy f or cr eat i ng geomet r y

We will put t he origin of t he coordinat e syst em at t he lower left corner of t he

rect angle. The coordinat es of t he corners are shown in t he figure below:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (5 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

We will first creat e four vert ices at t he four corners and j oin adj acent vert ices t o

get t he edges of t he rect angle. We will t hen form a face t hat covers t he area of

t he rect angle.

Cr eat e Ver t i ces

Find t he but t ons described below by point ing t he mouse at each of t he but t ons

and reading t he Descript ion Window.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ver t ex Com m a n d

Bu t t on > Cr ea t e Ver t ex

Not ice t hat t he Creat e Vert ex but t on has already been select ed by default .

Aft er you select a but t on under a sub- pad, it becomes t he default when you

go t o a different sub- pad and t hen come back t o t he sub- pad.

Creat e t he vert ex at t he lower- left corner of t he rect angle:

Next t o x : , ent er value 0. Next t o y : , ent er value 0. Next t o z: , ent er value 0

( t hese values should be default s) . Click Appl y . This creat es t he vert ex ( 0, 0, 0)

which is displayed in t he graphics window.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (6 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

I n t he Transcript window, GAMBI T report s t hat it "Creat ed vert ex: vert ex. 1".

The vert ices are numbered vert ex. 1, vert ex. 2 et c. in t he order in which t hey

are creat ed.

Repeat t his process t o creat e t hree more vert ices:

Vert ex 2: ( 0, 0. 1, 0)

Vert ex 3: ( 8, 0. 1, 0)

Vert ex 4: ( 8, 0, 0)

Not e t hat for a 2D problem, t he z- coordinat e can always be left t o t he default

value of 0.

Op er a t ion Toolp a d > Globa l Con t r ol > Fit t o Win d ow Bu t t on

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (7 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

This fit s t he four vert ices of t he rect angle we have creat ed t o t he size of t he

Graphics Window.

( Click pict ure for larger image)

Cr eat e Edges

We' ll now connect appropriat e pairs of vert ices t o form edges. To select any

ent it y in GAMBI T, hold down t he Shi f t key and click on t he ent it y.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Cr ea t e Ed ge

Select t wo vert ices t hat make up an edge of t his rect angle by holding down t he

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (8 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

Shi f t but t on and clicking on t he corresponding vert ices. As each vert ex is picked,

it will appear red in t he Graphics Window. Then let go of t he Shi f t but t on. We

can check t he select ed vert ices by clicking on t he up- arrow next t o Ver t i ces: .

This will bring up a window cont aining t he vert ices t hat have been select ed.

Vert ices can be moved from t he Avai l abl e and Pi ck ed list s by select ing t hem

and t hen pressing t he left or right arrow but t ons.

Aft er t he correct vert ices have been select ed, click Cl ose, t hen click Appl y in t he

Creat e St raight Edge window.

Repeat t his process t o creat e a rect angle.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (9 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

( Click pict ure for larger image)

Cr eat e Face

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on

> For m Fa ce

To form a face out of t he area enclosed by t he four lines, we need t o select t he

four ledges t hat enclose t his area. This can be done by holding down t he Shi f t

key, clicking on each line ( not ice t hat t he current ly select ed line appears red) ,

and t hen releasing t he Shi f t key aft er all four lines have been select ed.

Alt ernat ively, an easier way t o do t his would be t o click on t he up arrow next t o

edges:

This will bring up t he Edge List window. Click on Al l - > t o select all of t he edges

at once. Click Cl ose.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (10 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #1

Click Appl y t o creat e t he face.

Go t o St ep 2: Mesh Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step1.htm (11 of 11)11/7/2005 6:36:06 PM

Fluent Tutorial - Laminar Pipe Flow Step #2

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet r y i n GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 2: Mesh Geometry in GAMBIT

We' ll now creat e a mesh on t he rect angular face wit h 100 divisions in t he axial

direct ion and 5 divisions in t he radial direct ion. We' ll first mesh t he four edges

and t hen t he face. The desired grid spacing is specified t hrough t he edge mesh.

Mesh Edges

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on >

Mes h Ed ges

Shi f t - cl i ck or bring up t he Edge List window and select bot h t he vert ical lines. I f

t his is difficult , one can zoom in on an edge by holding down t he Ct r l but t on,

clicking and dragging t he mouse t o specify an area t o zoom in on, and releasing

t he Ct r l but t on. To ret urn t o t he main view, click on t he Globa l Con t r ol Toolp a d

> Fit t o Win d ow Bu t t on again.

Once a vert ical edge has been select ed, select I nt er val Count from t he drop

down box t hat says I nt er val Si ze in t he Mesh Edges Window. Then, in t he box

t o t he left of t his combo box, ent er 5 for t he int erval count .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step2.htm (1 of 3)11/7/2005 6:36:18 PM

Fluent Tutorial - Laminar Pipe Flow Step #2

Click Appl y . Nodes appear on t he edges showing t hat t hey are divided int o 5.

( Click pict ure for larger image)

Repeat t he same process for t he horizont al edges, but wit h an int erval count of

100.

Now t hat t he edges are meshed, we are ready t o creat e a 2- D mesh for t he face.

Mesh Face

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Mes h Fa ces

Shi f t l ef t - cl i ck on t he face or use t he up arrow next t o Faces t o select t he face.

Click Appl y .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step2.htm (2 of 3)11/7/2005 6:36:18 PM

Fluent Tutorial - Laminar Pipe Flow Step #2

( Click pict ure for larger image)

Go t o St ep 3: Specify Boundary Types in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step2.htm (3 of 3)11/7/2005 6:36:18 PM

Fluent Tutorial - Laminar Pipe Flow Step #3

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Speci f y Boundar y Ty pes i n GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 3: Specify Boundary Types in GAMBIT

Cr eat e Boundar y Ty pes

We' ll next set t he boundary t ypes in GAMBI T. The left edge is t he inlet of t he

pipe, t he right edge t he out let , t he t op edge t he wall, and t he bot t om edge t he

axis.

Op er a t ion Toolp a d > Zon es Com m a n d Bu t t on > Sp ecify Bou n d a r y Typ es

Com m a n d Bu t t on

This will bring up t he Specify Boundary Types window on t he Operat ion Panel.

We will first specify t hat t he left edge is t he inlet . Under Ent i t y : , pick Edges so

t hat GAMBI T knows we want t o pick an edge ( face is default ) .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step3.htm (1 of 4)11/7/2005 6:36:49 PM

Fluent Tutorial - Laminar Pipe Flow Step #3

Now select t he left edge by Shi f t - cl i ck i ng on it . The select ed edge should

appear in t he yellow box next t o t he Edges box you j ust worked wit h as well as

t he Label / Ty pe list right under t he Edges box.

Next t o Name: , ent er inlet.

For Ty pe: , select VELOCI TY_I NLET.

Click Appl y . You should see t he new ent ry appear under Name/ Ty pe box near

t he t op of t he window.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step3.htm (2 of 4)11/7/2005 6:36:49 PM

Fluent Tutorial - Laminar Pipe Flow Step #3

Repeat t his process for t he ot her t hree edges according t o t he following t able:

Edge

Position

Name Type

Left inlet VELOCITY_INLET

Right outlet PRESSURE_OUTLET

Top wall WALL

Bottom centerline AXIS

You should have t he following edges in t he Name/ Ty pe list when finished:

Sav e and Ex por t

Ma in Men u > File > Sa ve

Ma in Men u > File > Exp or t > Mes h . . .

Type in pipe.msh for t he Fi l e Name: . Select Ex por t 2d Mesh since t his is a 2

dimensional mesh. Click Accept .

Check pipe.msh has been creat ed in your working direct ory.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step3.htm (3 of 4)11/7/2005 6:36:49 PM

Fluent Tutorial - Laminar Pipe Flow Step #3

Go t o St ep 4: Set Up Problem in FLUENT

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step3.htm (4 of 4)11/7/2005 6:36:49 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Pr obl em i n FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 4: Set Up Problem in FLUENT

Launch Fl uent 6.0

St a r t > Pr ogr a m s > Flu en t I n c > FLUENT 6. 0

Select 2ddp from t he list of opt ions and click Run.

The "2ddp" opt ion is used t o select t he 2- dimensional, double- precision

solver. I n t he double- precision solver, each float ing point number is

represent ed using 64 bit s in cont rast t o t he single- precision solver which uses

32 bit s. The ext ra bit s increase not only t he precision but also t he range of

magnit udes t hat can be represent ed. The downside of using double precision

is t hat it requires more memory.

I mpor t Gr i d

Ma in Men u > File > Rea d > Ca s e. . .

Navigat e t o t he working direct ory and select t he pipe. msh file. This is t he mesh

file t hat was creat ed using t he preprocessor GAMBI T in t he previous st ep.

FLUENT report s t he mesh st at ist ics as it reads in t he mesh:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (1 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

Check t he number of nodes, faces ( of different t ypes) and cells. There are 500

quadrilat eral cells in t his case. This is what we expect since we used 5 divisions

in t he radial direct ion and 100 divisions in t he axial direct ion while generat ing t he

grid. So t he t ot al number of cells is 5* 100 = 500.

Also, t ake a look under zones. We can see t he four zones i nl et , out l et , w al l ,

and cent er l i ne t hat we defined in GAMBI T.

Check and Di spl ay Gr i d

First , we check t he grid t o make sure t hat t here are no errors.

Ma in Men u > Gr id > Ch eck

Any errors in t he grid would be report ed at t his t ime. Check t he out put and make

sure t hat t here are no errors report ed. Check t he grid size:

Ma in Men u > Gr id > I n fo > Size

The following st at ist ics should appear:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (2 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

Display t he grid:

Ma in Men u > Dis p la y > Gr id . . .

Make sure all 5 it ems under Sur f aces is select ed. Then click Di spl ay . The

graphics window opens and t he grid is displayed in it . You can now click Cl ose in

t he Grid Display menu t o get back some deskt op space. The graphics window will

remain.

Some of t he operat ions available in t he graphics window are:

Translat ion: The grid can be t ranslat ed in any direct ion by holding down t he Lef t

Mouse But t on and t hen moving t he mouse in t he desired direct ion.

Zoom I n: Hold down t he Mi ddl e Mouse But t on and drag a box from t he Upper

Lef t Hand Cor ner t o t he Low er Ri ght Hand Cor ner over t he area you want t o

zoom in on.

Zoom Out : Hold down t he Mi ddl e Mouse But t on and drag a box anywhere from

t he Low er Ri ght Hand Cor ner t o t he Upper Lef t Hand Cor ner .

Use t hese operat ions t o zoom int o t he grid t o obt ain t he view shown below.

Not e: The zooming operat ions cannot be performed wit hout a middle mouse

but t on.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (3 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

( Click pict ure for larger image)

You can also look at specific part s of t he grid by choosing t he boundaries you

wish t o view under Sur f aces ( click t o select and click again t o deselect a specific

boundary) . Click Di spl ay again when you have select ed your boundaries. For

example, t he w al l , out l et , and cent er l i ne boundaries have been select ed in t he

following view:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (4 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

These opt ions will display t he graph:

( Click pict ure for larger image)

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (5 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

For convenience, t he but t on next t o Sur f aces select s all of t he boundaries

while t he deselect s all of t he boundaries at once.

Close t he Grid Display Window when you are done.

Def i ne Sol ver Pr oper t i es

Ma in Men u > Defin e > Mod els > Solver

Choose Ax i sy mmet r i c under Space. We' ll use t he default s of segregat ed solver,

implicit formulat ion, st eady flow and absolut e velocit y formulat ion. Click OK.

Ma in Men u > Defin e > Mod els > Vis cou s

Lami nar flow is t he default . So we don' t need t o change anyt hing in t his menu.

Click Cancel .

Ma in Men u > Defin e > Mod els > En er gy

For incompressible flow, t he energy equat ion is decoupled from t he cont inuit y

and moment um equat ions. We need t o solve t he energy equat ion only if we are

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (6 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

int erest ed in det ermining t he t emperat ure dist ribut ion. We will not deal wit h

t emperat ure in t his example. So leave t he Ener gy Equat i on unselect ed and

click Cancel t o exit t he menu.

Def i ne Mat er i al Pr oper t i es

Ma in Men u > Defin e > Ma t er ia ls . . .

Change Densi t y t o 1.0 and Vi scosi t y t o 2e-3. These are t he values t hat we

specified under Problem Specificat ion. We' ll t ake bot h as const ant .

Click Change/ Cr eat e.

Def i ne Oper at i ng Condi t i ons

Ma in Men u > Defin e > Op er a t in g Con d it ion s . . .

For all flows, FLUENT uses gauge pressure int ernally. Any t ime an absolut e

pressure is needed, it is generat ed by adding t he operat ing pressure t o t he

gauge pressure. We' ll use t he default value of 1 at m ( 101, 325 Pa) as t he

Oper at i ng Pr essur e.

Click Cancel t o leave t he default in place.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (7 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

Def i ne Boundar y Condi t i ons

We' ll now set t he value of t he velocit y at t he inlet and pressure at t he out let .

Ma in Men u > Defin e > Bou n d a r y Con d it ion s . . .

We not e here t hat t he four t ypes of boundaries we defined are specified as zones

on t he left side of t he Boundary Condit ions Window. The cent er l i ne zone should

be select ed by default . Make sure it is, t hen make sure t he Ty pe of t his

boundary is select ed as ax i s and click Set .... Not ice t hat t here is not hing t o set

for t he axis. Click OK.

Move down t he list and select i nl et under Zone. Not e t hat FLUENT indicat es t hat

t he Ty pe of t his boundary is vel oci t y - i nl et . Recall t hat t he boundary t ype for

t he "inlet " was set in GAMBI T. I f necessary, we can change t he boundary t ype

set previously in GAMBI T in t his menu by select ing a different t ype from t he list

on t he right .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (8 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #4

Click on Set .... Ent er 1 for Vel oci t y Magni t ude. Click OK. This set s t he velocit y

of t he fluid ent ering at t he left boundary.

The ( absolut e) pressure at t he out let is 1 at m. Since t he operat ing pressure is

set t o 1 at m, t he out let gauge pressure = out let absolut e pressure - operat ing

pressure = 0. Choose out l et under Zone. The Ty pe of t his boundary is

pr essur e- out l et . Click on Set .... The default value of t he Gauge Pr essur e is 0.

Click Cancel t o leave t he default in place.

Last ly, click on w al l under Zones and make sure Ty pe is set as w al l . Click on

each of t he t abs and not e t hat only moment um can be changed under t he

current condit ions. This will not be so under lat er exercises so make a not e of t he

locat ion of t hese opt ions. Click OK.

Click Cl ose t o close t he Boundary Condit ions menu.

Go t o St ep 5: Solve!

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step4.htm (9 of 9)11/7/2005 6:37:26 PM

Fluent Tutorial - Laminar Pipe Flow Step #5

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Sol ve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 5: Solve!

We' ll use a second- order discret izat ion scheme.

Ma in Men u > Solve > Con t r ols > Solu t ion . . .

Change Moment um t o Second Or der Upw i nd.

Click OK.

Set I ni t i al Guess

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step5.htm (1 of 5)11/7/2005 6:38:25 PM

Fluent Tutorial - Laminar Pipe Flow Step #5

I nit ialize t he flow field t o t he values at t he inlet :

Ma in Men u > Solve > I n it ia lize > I n it ia lize. . .

I n t he Solut ion I nit ializat ion menu t hat comes up, choose i nl et under Comput e

Fr om. The Ax i al Vel oci t y for all cells will be set t o 1 m/ s, t he Radi al Vel oci t y

t o 0 m/ s and t he Gauge Pr essur e t o 0 Pa. These values have been t aken from

t he inlet boundary condit ion.

Click I ni t . This complet es t he init ializat ion.

Set Conver gence Cr i t er i a

FLUENT report s a residual for each governing equat ion being solved. The residual

is a measure of how well t he current solut ion sat isfies t he discret e form of each

governing equat ion. We' ll it erat e t he solut ion unt il t he residual for each equat ion

falls below 1e- 6.

Ma in Men u > Solve > Mon it or s > Res id u a l. . .

Change t he residual under Conver gence Cr i t er i on for cont i nui t y , x - vel oci t y ,

and y - vel oci t y , all t o 1e- 6.

Also, under Opt i ons, select Pl ot . This will plot t he residuals in t he graphics

window as t hey are calculat ed.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step5.htm (2 of 5)11/7/2005 6:38:25 PM

Fluent Tutorial - Laminar Pipe Flow Step #5

Click OK.

This complet es t he problem specificat ion. Save your work:

Ma in Men u > File > Wr it e > Ca s e. . .

Type in pipe.cas for Case Fi l e. Click OK. Check t hat t he file has been creat ed in

your working direct ory. I f you exit FLUENT now, you can ret rieve all your work at

any t ime by reading in t his case file.

I t er at e Unt i l Conver gence

St art t he calculat ion by running 100 it erat ions:

Ma in Men u > Solve > I t er a t e. . .

I n t he I t erat e Window t hat comes up, change t he Number of I t er at i ons t o 100.

Click I t er at e.

The residuals for each it erat ion is print ed out as well as plot t ed in t he graphics

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step5.htm (3 of 5)11/7/2005 6:38:25 PM

Fluent Tutorial - Laminar Pipe Flow Step #5

window as t hey are calculat ed.

( Click pict ure for larger image)

The residuals fall below t he specified convergence crit erion of 1e- 6 in 46

it erat ions.

Save t he solut ion t o a dat a file:

Ma in Men u > File > Wr it e > Da t a . . .

Ent er pipe.dat for Dat a Fi l e and click OK. Check t hat t he file has been creat ed

in your working direct ory. You can ret rieve t he current solut ion from t his dat a file

at any t ime.

Go t o St ep 6: Analyze Result s

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step5.htm (4 of 5)11/7/2005 6:38:25 PM

Fluent Tutorial - Laminar Pipe Flow Step #5

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step5.htm (5 of 5)11/7/2005 6:38:25 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Anal y ze Resul t s

7. Refine Mesh

Problem 1

Problem 2

Step 6: Analyze Results

Cent er l i ne Vel oci t y

We' ll plot t he variat ion of t he axial velocit y along t he cent erline.

Ma in Men u > Plot > XY Plot . . .

Make sure t hat Posi t i on on X Ax i s is set under Opt i ons, and X is set t o 1 and

Y t o 0 under Pl ot Di r ect i on. This t ells FLUENT t o plot t he x- coordinat e value on

t he abscissa of t he graph.

Under Y Ax i s Funct i on, pick Vel oci t y ... and t hen in t he box under t hat , pick

Ax i al Vel oci t y .

Please not e t hat X Ax i s Funct i on and Y Ax i s Funct i on describe t he x and y

axes of t he graph, which should not be confused wit h t he x and y direct ions of

t he pipe.

Finally, select cent er l i ne under Sur f aces since we are plot t ing t he axial velocit y

along t he cent erline. This finishes set t ing up t he plot t ing paramet ers.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (1 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Click Pl ot .

This brings up a plot of t he axial velocit y as a funct ion of t he dist ance along t he

cent erline of t he pipe.

( Click pict ure for larger image)

I n t he graph t hat comes up, we can see t hat t he velocit y reaches a const ant

value beyond a cert ain dist ance from t he inlet . This is t he fully- developed flow

region.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (2 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Change t he axes ext ent s: I n t he Solut ion XY Plot menu, click on Ax es.... Under

Opt i ons, deselect Aut o Range. The boxes under Range should now be

act ivat ed. Select X under Ax i s. Ent er 1 for Mi ni mum and 3 for Max i mum under

Range.

We' ll t urn on t he grid lines t o help est imat e where t he flow becomes fully

developed. Check t he boxes next t o Maj or Rul es and Mi nor Rul es under

Opt i ons. Click Appl y .

Now, pick Y under Ax i s and once again deselect Aut o Range under Opt i ons,

t hen ent er 1.8 for Mi ni mum and 2.0 for Max i mum under Range. Also select

Maj or Rul es and Mi nor Rul es t o t urn on t he grid lines in t he Y direct ion. We

have now finished specifying t he range for each axes, so click Appl y and t hen

Cl ose.

Go back t o t he Solut ion XY Plot menu and click Pl ot t o replot t he graph wit h t he

new axes ext ent s. We can see t hat t he fully- developed region st art s at around

x= 3m and t he cent erline velocit y in t his region is 1. 93 m/ s.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (3 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

( Click pict ure for larger image)

Sav i ng t he Pl ot

Save t he dat a from t his plot :

I n t he Solut ion XY Plot Window, check t he Wr i t e t o Fi l e box under Opt i ons. The

Pl ot but t on should have changed t o Wr i t e.... Click on Wr i t e.... Ent er vel.xy as

t he XY Fi l e Name and click OK. Check t hat t his file has been creat ed in your

FLUENT working direct ory.

Now, save a pict ure of t he plot :

Leave t he Solut ion XY Plot Window and t he Graphics Window open and click on:

File > Ha r d cop y . . .

Under For mat , choose one of t he following t hree opt ions:

EPS - if you have a post script viewer, t his is t he best choice. EPS allows you t o

save t he file in vect or mode, which will offer t he best viewable image qualit y.

Aft er select ing EPS, choose Vect or from under Fi l e Ty pe.

TI FF - t his will offer a high resolut ion image of your graph. However, t he image

file generat ed will be rat her large, so t his is not recommended if you do not have

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (4 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

a lot of room on your st orage device.

JPG - t his is small in size and viewable from all browsers. However, t he qualit y

of t he image is not part icularly good.

Aft er select ing your desired image format and associat ed opt ions, click on Save...

Ent er vel.eps, vel.tif, or vel.jpg depending on your format choice and click

OK.

Verify t hat t he image file has been creat ed in your working direct ory. You can

now copy t his file ont o a disk or print it out for your records.

Coef f i ci ent of Sk i n Fr i ct i on

FLUENT provides a large amount of useful informat ion in t he online help t hat

comes wit h t he soft ware. Let ' s probe t he online help for informat ion on

calculat ing t he coefficient of skin frict ion.

Ma in Men u > Help > Us er ' s Gu id e I n d ex. . .

Click on S in t he links on t op and scroll down t o sk i n f r i ct i on coef f i ci ent . Click

on t he second 965 link ( normally, you would have t o go t hrough each of t he

links unt il you find what you are looking for) . We can see an excerpt on t he skin

coefficient as well as t he equat ion for calculat ing it .

Click on t he link for Ref er ence Val ues panel , which t ells us how t o set t he

reference values used in calculat ing t he skin coefficient .

Set t he reference values:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (5 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Ma in Men u > Rep or t > Refer en ce Va lu es . . .

Select i nl et under Comput e Fr om t o t ell FLUENT t o calculat e t he reference

values from t he values at inlet . Check t hat densit y is 1 kg/ m3 and velocit y is 1 m/

s. ( Alt ernat ely, you could have j ust t yped in t he appropriat e values) . Click OK.

Go back t o t he Solut ion XY Plot menu. Uncheck Wr i t e t o Fi l e under Opt i ons

since we want t o plot t o t he window right now. We can leave t he ot her Opt i ons

and Pl ot Di r ect i on as is since we are st ill plot t ing against t he x dist ance along

t he pipe.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (6 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Under t he Y Ax i s Funct i on, pick Wal l Fl ux es..., and t hen Sk i n Fr i ct i on

Coef f i ci ent in t he box under t hat .

Under Sur f aces, select w al l and unselect cent er l i ne by clicking on t hem.

Reset axes ranges: Go t o Ax es... and re- select Aut o- Range for t he Y axis. Set

t he range of t he X axis from 1 t o 8 by select ing X under Ax i s, ent ering 1 under

Mi ni mum, and 8 under Max i mum in t he Range box ( remember t o de- select

Aut o- Range first if it is checked) .

Click Appl y , Cl ose, and t hen Pl ot in t he Solut ion XY Plot Window.

( Click pict ure for larger image)

We can see t hat t he fully developed region is reached at around x= 3. 0m and t he

skin frict ion coefficient in t his region is around 1. 54. Compare t he numerical

value of 1. 54 wit h t he t heoret ical, fully- developed value of 0. 16.

Save t he dat a from t his plot : Pick Wr i t e t o Fi l e under Opt i ons and click

Wr i t e.... Ent er cf.xy for XY Fi l e and click OK.

Vel oci t y Pr of i l e

We' ll next plot t he velocit y at t he out let as a funct ion of t he dist ance from t he

cent er of t he pipe. To do t his, we have t o set t he y axis of t he graph t o be t he y

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (7 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

axis of t he pipe ( t he radial direct ion) .

To plot t he posit ion variable on t he y axis of t he graph, uncheck Posi t i on on X

Ax i s under Opt i ons and choose Posi t i on on Y Ax i s inst ead. To make t he

posit ion variable t he radial dist ance from t he cent erline, under Pl ot Di r ect i on,

change X t o 0 and Y t o 1. To plot t he axial velocit y on t he x axis of t he graph, for

X Ax i s Funct i on, pick Vel oci t y ... and Ax i al Vel oci t y under t hat .

Since we want t o plot t his at t he out let boundary, pick out l et under Sur f aces.

Change bot h t he x and y axes t o Aut o- Range.

Uncheck Wr i t e t o Fi l e under Opt i ons so t hat we can see t he graph. Click Pl ot .

( Click pict ure for larger image)

Does t his look like a parabolic profile?

Save t he dat a from t his plot : Pick Wr i t e t o Fi l e under Opt i ons and click

Wr i t e.... Ent er profile.xy for XY Fi l e and click OK.

To see how t he velocit y profile changes in t he developing region, let us add t he

profiles at x= 0. 6m ( x/ D= 3) and x= 0. 12m ( x/ D= 6) t o t he above plot . First ,

creat e a line at x= 0. 6m using t he Line/ Rake t ool:

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (8 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Ma in Men u > Su r fa ce > Lin e/ Ra ke

We' ll creat e a st raight line from ( x0, y0) = ( 0. 6, 0) t o ( x1, y1) = ( 0. 6, 0. 1) . Select

Li ne Tool under Opt i ons. Ent er x 0= 0.6, y 0= 0, x 1= 0.6, y 1= 0.1. Ent er line1

under New Sur f ace Name. Click Cr eat e.

To see t he line j ust creat ed, select

Ma in Men u > Dis p la y > Gr id . . .

Not e t hat l i ne1 appears in t he list of surfaces. Select all surfaces except def aul t -

i nt er i or . Click Di spl ay . This displays all surfaces but not t he mesh cells. Zoom

int o t he region near t he inlet t o see t he line creat ed at x= 0. 6m. ( Click here t o

review t he zoom funct ionalit y discussion in st ep 4. ) line1 is t he whit e vert ical line

t o t he right in t he figure below.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (9 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

Similarly, creat e a vert ical line called line2 at x= 1. 2; ( x0, y0) = ( 1. 2, 0) t o ( x1, y1) =

( 1. 2, 0. 1) in t his case. Display it in t he graphics window t o check t hat it has been

creat ed correct ly.

Now we can plot t he velocit y profiles at x= 0. 6m ( x/ D= 3) and x= 0. 12m ( x/ D= 6)

along wit h t he out let profile. I n t he Solut ion XY plot menu, use t he same set t ings

as above. Under Sur f aces, in addit ion t o out l et , select l i ne1 and l i ne2. Select

Node Val ues under Opt ions. Click Pl ot . Your symbols might be different from

t he ones below. You can change t he symbols and line st yles under t he Cur ves...

but t on. Click on Hel p in t he Curves menu if you have problems figuring out how

t o change t hese set t ings.

The profile t hree diamet ers downst ream is fairly close t o t he fully- developed

profile at t he out let . I f you redo t his plot using t he fine grid result s in t he next

st ep, you' ll see t hat t his is not act ually t he case. The coarse grid used here

doesn' t capt ure t he boundary layer development properly and underpredict s t he

development lengt h.

I n FLUENT, you can choose t o display t he comput ed cell- cent er values or

values t hat have been int erpolat ed t o t he nodes. By default , t he Node Values

opt ion is t urned on, and t he int erpolat ed values are displayed. Node- averaged

dat a curves may be somewhat smoot her t han curves for cell values.

Vel oci t y Vect or s

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (10 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

One can plot vect ors in t he ent ire domain, or on select ed surfaces. Let us plot

t he velocit y vect ors for t he ent ire domain t o see how t he flow develops

downst ream of t he inlet .

Ma in Men u > Dis p la y > Vect or s . . . > Dis p la y

Zoom int o t he region near t he inlet . ( Click here t o review t he zoom funct ionalit y

discussion in st ep 4. ) The lengt h and color of t he arrows represent t he velocit y

magnit ude. The vect or display is more int elligible if one makes t he arrows

short er as follows: Change Scal e t o 0.4 in t he Vect ors menu and click Di spl ay .

You can reflect t he plot about t he axis t o get an expanded sect ional view:

Mai n Menu > Di spl ay > Vi ew s...

Under Mi r r or Pl anes, only t he ax i s surface is list ed since t hat is t he only

symmet ry boundary in t he present case. Select ax i s and click Appl y . Close t he

Views window.

The velocit y vect ors provide a pict ure of how t he flow develops downst ream of

t he inlet . As t he boundary layer grows, t he flow near t he wall is ret arded by

viscous frict ion. Not e t he sloping arrows in t he near wall region close t o t he inlet .

This indicat es t hat t he slowing of t he flow in t he near- wall region result s in an

inj ect ion of fluid int o t he region away from t he wall t o sat isfy mass conservat ion.

Thus, t he velocit y out side t he boundary layer increases.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (11 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #6

By default , one vect or is drawn at t he cent er of each cell. This can be seen by

t urning on t he grid in t he vect or plot : Select Dr aw Gr i d in t he Vect ors menu

and t hen click Di spl ay in t he Grid Display as well as t he Vect ors menus.

Velocit y vect ors are t he default , but you can also plot ot her vect or quant it ies.

See sect ion 27. 1. 3 of t he user manual for more det ails about t he vect or plot

funct ionalit y.

Go t o St ep 7: Refine Mesh

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step6.htm (12 of 12)11/7/2005 6:39:21 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Lami nar Pi pe Fl ow

Problem Specificat ion

1. St art - up and preliminary set - up

2. Creat e Geomet ry

3. Mesh Geomet ry

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Ref i ne Mesh

Problem 1

Problem 2

Step 7: Refine Mesh

I t is very import ant t o assess t he dependence of your result s on t he mesh used

by repeat ing t he same calculat ion on different meshes and comparing t he

result s. We will re- do t he previous calculat ion on a 100x10 mesh and compare

t he result s wit h t he 100x5 mesh used previously. I f you prefer t o skip t he

GAMBI T st eps for modifying t he mesh, download t he 100x10 mesh ( by right -

clicking on t he link) and go direct ly t o t he FLUENT analysis discussed below.

Modi f y Mesh i n GAMBI T

The 100x5 mesh is saved as pipe. dbs in your working direct ory. Bring up t he

command prompt window as in st ep 1. To copy pipe. dbs t o pipe2. dbs, at t he

command prompt , t ype

copy pipe.dbs pipe2.dbs

We will work wit h pipe2. dbs in order t o ret ain pipe. dbs as is. Launch GAMBI T

wit h pipe2. dbs as t he input file by t yping:

gambit pipe2.dbs

Not e in t he main menu bar t hat pipe2 is t he I D of t his j ob. So files creat ed during

t his session will have t hat prefix.

We will delet e t he face mesh, modify t he edge meshes for t he vert ical edges and

remesh t he face. To delet e t he original face mesh, choose

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on > Delet e

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (1 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Fa ce Mes h es

I n t he Delet e Face Meshes Window t hat comes up, uncheck t he Remove

unused l ow er mesh box. This t ells GAMBI T t o remove t he face mesh only and

keep t he edge meshes associat ed wit h t he face mesh. Since we will be changing

t he mesh on only t wo edges of t he rect angle, t here is no need t o redo t he

meshes for all four edges.

Select t he only face of t he rect angle by shi f t - cl i ck i ng on it and t hen click Appl y .

Modi f y Edge Meshes

To change t he number of divisions on t he vert ical edges from 5 t o 10, choose:

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on > Mes h

Ed ges

Select t he t wo vert ical edges by holding down t he Shi f t but t on, clicking on each

in t urn, and t hen releasing t he Shi f t but t on. Select I nt er val count from t he box

under Spaci ng t hat says I nt er val si ze. Change t he number in t he box next t o

t he I nt er val count box from 5 t o 10.

Make sure t hat t he Remove ol d mesh box is checked under Opt i ons. This will

make sure t hat t he old edge meshes are erased before t he new edge meshes are

creat ed.

Click Appl y .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (2 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Remember t hat you can zoom in by holding down Ct r l , dragging a box across

t he area you want t o zoom in on, and t hen releasing Ct r l . Do t his now and make

sure t hat t he vert ical edges have 10 divisions.

( Click image for larger pict ure)

Recr eat e Face Mesh

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on > Mes h

Fa ces

Shi f t - cl i ck on t he face in t he Graphics Window t o select it . Click Appl y .

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (3 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( Click here for larger pict ure)

Sav e & Ex por t

Ma in Men u > File > Sa ve

Ma in Men u > File > Exp or t > Mes h . . .

Type in pipe2.msh for t he Fi l e Name: . Select Ex por t 2d Mesh opt ion. Click

Accept .

Fi ner Mesh Anal y si s

Repeat st eps 4 and 5 of t his t ut orial wit h t he 100x10 mesh ( a t ad on t he

repet it ious side but consider it good pract ice) .

One you obt ain t he solut ion, plot t he variat ion of t he cent erline velocit y along t he

x- direct ion as described in st ep 6. Compare t his result wit h t hat obt ained on t he

previous mesh which is st ored in t he vel. xy file creat ed earlier. To do t his, aft er

cent erline velocit y has been plot t ed, click on Load Fi l e... in t he Solut ion XY Plot

window. Navigat e t o your working folder if necessary and click on vel.xy and

OK. Click Pl ot .

I n t he graphics window, we can see bot h of t he lines plot t ed in t he same window.

Adj ust t he axes so t hat you can zoom in on t he beginning of t he fully developed

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (4 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

region.

( Click image for larger pict ure)

I n t he cent erline velocit y plot above, t he whit e and red symbols represent t he

result s on t he 100x10 mesh and 100x5 meshes, respect ively. The cent erline

velocit y in t he fully- developed region for t he finer mesh is 1. 98 m/ s. This value

agrees bet t er wit h t he analyt ical value of 2 m/ s t hat t he value of 1. 93 m/ s

obt ained on t he coarser mesh. Save t he dat a for t his plot as vel2.xy. The

velocit y result get s more accurat e on refining t he mesh as expect ed.

Plot t he skin frict ion coefficient as described in st ep 6. Compare t he result wit h

t hat obt ained on t he 100x5 mesh by loading it from cf.xy.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (5 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( Click here for larger image)

The finer mesh provides a skin frict ion coefficient of 0. 159 in t he fully- developed

region, which is much closer t o t he t heoret ical value of 0. 16 t han t he

corresponding coarser mesh value of 0. 154. Save t he dat a for t his plot as cf2.

xy.

Similarly, plot t he velcoit y profile at t he out let and compare wit h t he coarser grid

result in out.xy. The t wo result s compare well wit h t he great est deviat ion

occurring near t he cent erline. Save t he dat a for t his plot as out2.xy.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (6 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( Click pict ure for larger image)

I f you repeat t he calculat ion on a 100x20 mesh, you' ll see t hat t he result s on t he

t wo finest meshes are grid- independent t o a high level of accuracy. I n t he plot s

below, t he whit e, red and green symbols correspond t o t he 100x20, 100x10 and

100x5 meshes, respect ively.

Vel oci t y al ong cent er l i ne:

( Click pict ure for larger image)

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (7 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Sk i n Coef f i ci ent :

( Click pict ure for larger image)

Out l et Vel oci t y :

( Click pict ure for larger image)

Go t o Problem 1

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (8 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/step7.htm (9 of 9)11/7/2005 6:40:22 PM

Fluent Tutorial - Problem #1

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Pr obl em 1

Problem 2

Problem 1

Pr obl em

a) Consider t he problem solved in t his t ut orial. At t he exit of t he pipe, we can

define t he error in t he calculat ion of t he cent erline velocit y as:

where Uc is t he cent erline value from FLUENT and Uexact is t he exact analyt ical

value for fully- developed laminar pipe flow. We expect t he error t o t ake t he form:

where t he coefficient K and t he power p depend upon t he met hod . Consider t he

solut ions obt ained on t he 100x5, 100x10, and 100x20 meshes. Using MATLAB,

perform a linear least squares fit of:

t o obt ain t he coefficient s K and p. You can look up t he value of Uexact from any

int roduct ory t ext book in fluid mechanics such as Fluid Mechanics by F. Whit e.

Explain why your values make sense.

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/ps1.htm (1 of 2)11/7/2005 6:40:29 PM

Fluent Tutorial - Problem #1

b) Repeat t he above exercise using t he "first - order upwind" scheme for t he

moment um equat ion. Cont rast t he value of p obt ained in t his case wit h t he

previous one and explain your result s briefly ( 2- 3 sent ences) .

Hi nt s

Not e t hat t he first or second order discret izat ion applies only t o t he convect ive

t erms in t he Navier- St okes equat ions. The viscous t erms are always second

order accurat e.

Go t o Problem 2

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/ps1.htm (2 of 2)11/7/2005 6:40:29 PM

Fluent Tutorial - Problem #2

Lami nar Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Pr obl em 2

Problem 2

Pr obl em

On your finest mesh ( 100x20) , rerun t he FLUENT calculat ion for Reynolds

numbers 200 and 500 using t he "second- order upwind" scheme. Not e: change

t he Reynolds number by adj ust ing t he molecular viscosit y µ. Plot t he cent erline

velocit y and skin frict ion as a funct ion of axial dist ance for Re = 100 ( previous

problem) , 200, and 500. Plot all t hree cases on t he same graph for comparsion.

Briefly explain t he t rend you observe as t he Reynolds number increases.

Hi nt s

I f you' ve saved t he 100x20 mesh in st ep 7, you can load it int o FLUENT again

wit hout having t o recreat e it in GAMBI T.

Solve for µ for each of t he Reynolds number first and t hen t hink about what

st eps need t o be changed.

Sol ut i on

Your solut ion should look somet hing like t he plot s below:

Cent er l i ne Vel oci t y

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/ps2.htm (1 of 3)11/7/2005 6:41:05 PM

Fluent Tutorial - Problem #2

( Click pict ure for larger image)

Sk i n Coef f i ci ent

( Click pict ure for larger image)

Back t o Problem Specificat ion

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/ps2.htm (2 of 3)11/7/2005 6:41:05 PM

Fluent Tutorial - Problem #2

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe1/ps2.htm (3 of 3)11/7/2005 6:41:05 PM

Fluent Tutorial: Turbulent Pipe Flow

Fl uent 6.0: Tur bul ent Pi pe Fl ow

Aut hor: Raj esh Bhaskaran

E- mail: rb88@cornell. edu

Pr obl em Speci f i cat i on

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem Specification

Let ' s revisit t he pipe flow example considered in t he previous exercise. As before,

t he inlet velocit y is 1 m/ s, t he fluid exhaust s int o t he ambient at mosphere and

densit y is 1 kg/ m

3

. For µ = 2 x 10

- 5

kg/ ( ms) , t he Reynolds no. based on t he pipe

diamet er and average velocit y at t he inlet is

At t his Reynolds number, t he flow is usually complet ely t urbulent .

A t urbulent flow exhibit s small- scale fluct uat ions in t ime. I t is usually not

possible t o resolve t hese fluct uat ions in a CFD calculat ion. So t he flow variables

such as velocit y, pressure, et c. are t ime- averaged. Unfort unat ely, t he t ime-

averaged governing equat ions are not closed i. e. t hey cont ain fluct uat ing

quant it ies which need t o be modeled using a t urbulence model. No t urbulence

model is current ly available t hat is valid for all t ypes of flows and so it is

necessary t o choose and fine- t une a model for part icular classes of flows. I n t his

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/index.htm (1 of 2)11/7/2005 6:41:13 PM

Fluent Tutorial: Turbulent Pipe Flow

exercise, you' ll be t urned loose on variant s of t he k- ε model. But in t he real

world, t read wit h great caut ion: you should evaluat e t he validit y of your

calculat ions using a t urbulence model very carefully ( which, ahem, means t hat

t here is no get t ing away from st udying fluid dynamics concept s and numerical

met hods very carefully) . FLUENT should not be used as a black box. The k- ε

models consist of t wo different ial equat ions: one each for t he t urbulent kinet ic

energy k and t urbulent dissipat ion ε. These t wo equat ions have t o be solved

along wit h t he t ime- averaged cont inuit y, moment um and energy equat ions. So

t urbulent flow calculat ions are much more difficult and t ime- consuming t han

laminar flow calculat ions. This is an exercise t o whet your appet it e for t urbulent

flow calculat ions.

Go t o St ep 1: Creat e Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/index.htm (2 of 2)11/7/2005 6:41:13 PM

Compressible Flow in Nozzle - Problem Specification

Compr essi bl e Fl ow i n a Nozzl e

Aut hor: Raj esh Bhaskaran

E- mail: rb88@cornell. edu

Pr obl em Speci f i cat i on

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

* * Under const r uct i on* *

Problem Specification

Consider air flowing at high- speed t hrough a convergent - divergent nozzle having

a circular cross- sect ional area, A, t hat varies wit h axial dist ance from t he t hroat ,

x, according t o t he formula

A = 0. 1 + x

2

; - 0. 5 < x < 0. 5

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/index.htm (1 of 2)11/7/2005 6:41:22 PM

Compressible Flow in Nozzle - Problem Specification

where A is in square met ers and x is in met ers. The st agnat ion pressure p

o

at t he

inlet is 101, 325 Pa. The st agnat ion t emperat ure T

o

at t he inlet is 300 K. The

st at ic pressure p at t he exit is 3, 738. 9 Pa. We will calculat e t he Mach number,

pressure and t emperat ure dist ribut ion in t he nozzle using FLUENT and compare

t he solut ion t o quasi- 1D nozzle flow result s. The Reynolds number for t his high-

speed flow is large. So we expect viscous effect s t o be confined t o a small region

close t o t he wall. So it is reasonable t o model t he flow as inviscid.

Go t o St ep 1: Creat e Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/index.htm (2 of 2)11/7/2005 6:41:22 PM

Flow over an Airfoil - Problem Specification

Fl ow over an Ai r f oi l

Aut hor: Raj esh Bhaskaran

E- mail: rb88@cornell. edu

Pr obl em Speci f i cat i on

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Problem Specification

Consider air flowing over t he given airfoil. The freest ream velocit y is 50 m/ s and

t he angle of at t ack is 5

o

. Assume st andard sea- level values for t he freest ream

propert ies:

Pressure = 101, 325 Pa

Densit y = 1. 2250 kg/ m3

Temperat ure = 288. 16 K

Kinemat ic viscosit y v = 1. 4607e- 5 m

2

/ s

Det ermine t he lift and drag coefficient s under t hese condit ions using FLUENT.

Go t o St ep 1: Creat e Geomet ry in GAMBI T

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/index.htm (1 of 2)11/7/2005 6:41:32 PM

Flow over an Airfoil - Problem Specification

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/index.htm (2 of 2)11/7/2005 6:41:32 PM

Fluent Tutorial - Forced Convection on a Flat Plate

For ced Convect i on over a Fl at Pl at e

Aut hor: Mat t hew Offerman

E- mail: mvo2@cornell. edu

Pr obl em Speci f i cat i on

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem Specification

I n our problem, we have a flat plat e at a const ant t emperat ure of 413K. The

plat e is infinit ely wide. The velocit y profile of t he fluid is uniform at t he point x =

0. The free st ream t emperat ure of t he fluid is 353K. The assumpt ion of

incompressible flow becomes invalid increasingly less valid for larger

t emperat ure differences bet ween t he plat e and freest ream. Because of t his, we

will t reat t his as a compressible flow. We will analyze a fluid flow wit h t he

following non- dimensional condit ions:

http://instruct1.cit.cornell.edu/courses/fluent/plate/index.htm (1 of 3)11/7/2005 6:41:54 PM

Fluent Tutorial - Forced Convection on a Flat Plate

I n order t o achieve t hese flow condit ions, we will use t hese free st ream flow

condit ions:

According t o t he ideal gas law, t his t emperat ure and pressure result in t he

following freest ream densit y:

These flow condit ions do not necessarily represent a realist ic fluid. Rat her, t hey

are chosen t o provide t he Prandt l and Reynolds numbers specified above. This

will make calculat ions simpler t hroughout t his t ut orial.

Solve t his problem in FLUENT. Validat e t he solut ion by plot t ing t he y+ values at

t he plat e. Also plot t he velocit y profile at x = 1m. Then plot Reynolds Number vs.

Nusselt Number. Compare t he accuracy of your result s from FLUENT wit h

empirical correlat ions.

Preliminary Analysis

We expect t he t urbulent boundary layer t o grow along t he plat e. As t he boundary

layer grows in t hickness, t he rat e of heat t ransfer ( q' ' ) and t hus t he heat t ransfer

coefficient ( h) will decrease.

http://instruct1.cit.cornell.edu/courses/fluent/plate/index.htm (2 of 3)11/7/2005 6:41:54 PM

Fluent Tutorial - Forced Convection on a Flat Plate

We will compare t he numerical result s wit h experiment ally- derived heat t ransfer

correlat ions. We will creat e t he geomet ry and mesh in GAMBI T, read t he mesh

int o FLUENT, and solve t he flow problem.

Go t o St ep 1: Creat e Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/index.htm (3 of 3)11/7/2005 6:41:54 PM

Fluent Tutorial - Turbulent Pipe Flow Step #1

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Cr eat e Geomet r y i n GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Step 1: Create Geometry in GAMBIT

I f you would prefer t o skip t he mesh creat ion st eps, you can download t he

mesh here ( right click and select Save As...) and go st raight t o st ep 4.

Since t he flow is axisymmet ric, t he geomet ry is a rect angle as in t he Laminar

Pipe Flow t ut orial. We will first use a 100x30 mesh ( i. e. 100 divisions in t he axial

direct ion and 30 divisions in t he radial direct ion) .

We could creat e t his mesh from scrat ch, as in t he Laminar Pipe Flow t ut orial, but

inst ead, we will modify t he previous 100x5 t o get t he 100x30 mesh. This will

int roduce you t o t he art of modifying meshes in GAMBI T.

Go t o St ep 2: Mesh Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step1.htm11/7/2005 6:47:41 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet r y i n GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Step 2: Mesh Geometry in GAMBIT

Launch GAMBI T

Creat e a folder called pipe2 at a convenient locat ion t o use as your working

folder. Copy your pipe. dbs file cont aining t he 100x5 mesh from t he Laminar Pipe

Flow t ut orial t o t his folder. I f you don' t have t his file, here' s a copy ( r i ght - cl i ck

and select Save As...) . Rename t his file as pipe100x30. dbs. We' ll modify t his file

t o obt ain t he mesh for t he t urbulent pipe flow simulat ion.

St art GAMBI T in your working folder by t yping gambit -id pipe100x30 at t he

command prompt . ( Refer t o st ep 1 of t he Laminar Pipe Flow t ut orial if you' ve

forgot t en how t o do t his. ) Recall t hat GAMBI T will use t he id pipe100x30 as t he

default prefix for all files creat ed during t his session.

To make best use of screen real est at e, resize t he GAMBI T and browser windows

so t hat you approximat e t his screen arrangement . This way you can read

inst ruct ions in t he browser window and implement t hem in GAMBI T.

The mesh from t he previous t ut orial should be displayed. To fit t he mesh t o t he

size of t he window, select :

Globa l Con t r ol > Fit t o Win d ow

Del et e Pr evi ous Face Mesh

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (1 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

The first st ep we have t o do is remove t he old face mesh. Recall t hat t he face

mesh is built on t op of meshed edges, t hereby forming t he grid. I n t his case, we

don' t want t o remove t he underlying edge meshes. So t o delet e only t he face

mesh, select :

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Delet e Fa ce Mes h es

Since we only have one face, shi f t - cl i ck any edge of t he bounding rect angle t o

select t he face mesh we want t o delet e. The face you have select ed should

become red and t he name of t he face list ed in t he Delet e Face Meshes window in

t he drop down box.

Now, because we don' t want t o delet e t he edge meshes, uncheck t he Remove

unused l ow er mesh box.

Click Appl y .

Check t hat t he face mesh has been removed in t he GAMBI T Graphics Window.

Remesh Edges

Since we are st ill going t o use 100 divisions for t he horizont al edges, we only

need t o remesh t he vert ical edges.

To resolve t he much higher gradient near t he wall for a t urbulent flow, we will

use smaller grid spacing near t he wall by employing grid st ret ching.

For each vert ical edge, we will specify t he division lengt h next t o t he wall t o be

0. 001 and t he t ot al number of divisions t o be 30. I n GAMBI T, each edge has a

direct ion associat ed wit h it as shown by an arrow. We will set t his arrow t o point

away from t he wall. Then t he division next t o t he wall becomes t he "First Lengt h"

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (2 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

and t he division next t o t he axis becomes t he "Last Lengt h". We' ll specify t he

"First Lengt h" t o be 0. 001 and t he t ot al number of divisions t o be 30 for t he

edge; GAMBI T will aut omat ically calculat e t he appropriat e value for t he "Last

Lengt h".

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

Select t he vert ical edges by shift - clicking on each of t hem. Not ice t he red arrow

t hat appears on t he edge when it is select ed. Make sure t hese arrows are

point ing down ( t owards t he axis and away from t he wall) . I f bot h of t hese arrows

are point ing in t he wrong direct ion, you can reverse t hem by clicking Rever se

next t o Pi ck w i t h l i nk s. However, if only one of t he edges needs t o be reversed,

you can do t hat by shift - middle clicking on t hat edge. You' ll have t o zoom in t o

be able t o do t his. ( Recall t hat you can zoom in by holding down t he Ct rl key and

t hen dragging a box wit h your left mouse but t on. Double- click wit h t he middle

mouse but t on t o go back t o t he last view. )

For Ty pe in t he Mesh Edges menu, select Fi r st Lengt h from t he drop down box.

Next t o Lengt h, t ype in 0.001.

We want 30 divisions on each of t he vert ical edges; so select I nt er val Count

from t he drop down box under Spaci ng and ent er 30 in t he t ext box t o it s left .

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (3 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

Click Appl y .

I f you zoom in on t he right edge, you should see t he following:

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (4 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

( Click pict ure for larger image)

Not e t hat t he mesh spacing is smaller near t he wall as indicat ed by t he blue

circles on t he edge.

Recr eat e Face Mesh

The next st ep is t o recreat e t he face mesh on t op of t hese edge meshes. This is

t he same procedure as in t he previous t ut orial:

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Mes h Fa ces

Shi f t l ef t - cl i ck on t he face and click Appl y . The meshed area should look like

t his aft er zooming in:

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (5 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turbulent Pipe Flow Step #2

( Click pict ure for larger image)

Go t o St ep 3: Specify Boundary Types in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step2.htm (6 of 6)11/7/2005 6:48:48 PM

Fluent Tutorial - Turblent Pipe Flow Step #3

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Speci f y Boundar y Ty pes i n GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Step 3: Specify Boundary Types in GAMBIT

Recall t hat we creat ed t he following boundary t ypes for t he 100x5 mesh in t he

Laminar Pipe Flow t ut orial:

Edge

Position

Name Type

Left inlet VELOCITY_INLET

Right outlet PRESSURE_OUTLET

Top wall WALL

Bottom centerline AXIS

These boundary t ypes are st ill ret ained even if t he edges are remeshed since t he

edges t hemselves were not delet ed. To verify t his:

Op er a t ion Toolp a d > Zon es Com m a n d Bu t t on > Sp ecify Bou n d a r y Typ es

Check t hat t he following is in t he Name/ Ty pe list :

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step3.htm (1 of 2)11/7/2005 6:48:53 PM

Fluent Tutorial - Turblent Pipe Flow Step #3

Addit ionally, click on show l abel s. You should now be able t o see each of t he

boundary names on t he respect ive edges in t he Graphics Window. Verify t hat t he

boundary t ypes specificat ion is correct .

Sav e and Ex por t

As in t he previous t ut orial, we will now save and export t he mesh.

Ma in Men u > File > Sa ve

Ma in Men u > File > Exp or t > Mes h . . .

Type in pipe100x30.msh for t he Fi l e Name: . Select Ex por t 2d Mesh since t his

is a t wo- dimensional mesh. Click Accept .

Check t hat pipe100x30.msh has been creat ed in your working direct ory.

Exit GAMBI T: Ma in Men u > File > Exit and save t he session.

Go t o St ep 4: Set Up Problem in Fluent

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step3.htm (2 of 2)11/7/2005 6:48:53 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Pr obl em i n FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Step 4: Set Up Problem in FLUENT

Launch FLUENT

St a r t > Pr ogr a m s > Flu en t I n c > FLUENT 6. 0

Select 2ddp ( 2D, double- precision version) from t he list of opt ions and click

Run.

I mpor t Fi l e

Ma in Men u > File > Rea d > Ca s e. . .

Navigat e t o your working direct ory and select t he pipe100x30.msh file. Click OK.

The following should appear in t he FLUENT window:

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (1 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Check t he number of nodes, faces ( of different t ypes) and cells. There are 3000

quadrilat eral cells in t his case. This is what we' d expect since we used 30

divisions in t he radial direct ion and 100 divisions in t he axial direct ion while

generat ing t he grid. So t he t ot al number of cells is 30* 100 = 3000.

Also, t ake a look under zones. We can see t he four zones i nl et , out l et , w al l ,

and cent er l i ne t hat we defined in GAMBI T.

Gr i d

First , we check t he grid t o make sure t hat t here are no errors.

Ma in Men u > Gr id > Ch eck

Any errors in t he grid would be report ed at t his t ime. Check t he out put and make

sure t hat t here are no errors report ed. Then select :

Ma in Men u > Gr id > I n fo > Size

The following summary about t he grid should appear:

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (2 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Let ' s look at t he grid:

Ma in Men u > Dis p la y > Gr id . . .

Make sure all 5 it ems under Sur f aces are select ed. Then click Di spl ay .

Remember t hat we can zoom in using t he middle mouse but t on. Zoom in and

admire t he grid. How many divisions are t here in t he radial direct ion?

( Click pict ure for larger image)

Recall t hat you can look at specific component s of t he grid by choosing t he

ent it ies you wish t o view under Sur f aces ( click t o select and click again t o

deselect a specific boundary) . Click Di spl ay again when you have select ed your

boundaries. Use t his feat ure and make sure t hat t he boundary labels correspond

t o t he correct geomet ric ent it ies.

Close t he Grid Display Window when you are done.

Def i ne Sol ver Pr oper t i es

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (3 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Ma in Men u > Defin e > Mod els > Solver

Choose Ax i sy mmet r i c under Space. As in t he laminar pipe flow t ut orial, we' ll

use t he default s of segregat ed solver, implicit formulat ion, st eady flow and

absolut e velocit y formulat ion. Click OK.

Ma in Men u > Defin e > Mod els > Vis cou s . . .

Choose k - epsi l on ( 2eqn) . Not ice t hat t he window expands and addit ional

opt ions are displayed on choosing t he k- epsilon t urbulence model. Under Near -

Wal l Tr eat ment , pick Enhanced Wal l Tr eat ment so t hat we may get a more

accurat e result .

Click OK.

Ma in Men u > Defin e > Mod els > En er gy. . .

The energy equat ion can be t urned off since t his is an incompressible flow and

we are not int erest ed in t he t emperat ure. Make sure no t ick mark appears next

t o Ener gy Equat i on.

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (4 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Ma in Men u > Defin e > Ma t er ia ls . . .

Change Densi t y t o 1.0 and Vi scosi t y t o 2e-5. These are t he values in t he

Problem Specificat ion. We' ll t ake bot h as const ant .

Click Change/ Cr eat e.

Def i ne Oper at i ng Condi t i ons

Ma in Men u > Defin e > Op er a t in g Con d it ion s . . .

Recall t hat for all flows, FLUENT uses t he gauge pressure int ernally. Any t ime an

absolut e pressure is needed, it is generat ed by adding t he operat ing pressure t o

t he gauge pressure. We' ll use t he default value of 1 at m ( 101, 325 Pa) as t he

Oper at i ng Pr essur e.

Click Cancel t o leave t he default in place.

Def i ne Boundar y Condi t i ons

We' ll now set t he value of t he velocit y at t he inlet and pressure at t he out let .

Ma in Men u > Defin e > Bou n d a r y Con d it ion s . . .

The four t ypes of boundaries we defined are specified as zones on t he left side of

t he Boundary Condit ions Window. Recall t hat we don' t need t o set any

paramet ers for t he cent er l i ne and w al l zones. Verify t his by select ing each of

t hese t wo zones and clicking on Set ....

Choose inlet and click on Set .... Ent er 1 for Vel oci t y Magni t ude. This indicat es

t hat t he fluid is coming in normal t o t he inlet at t he rat e of 1 met er per second.

Select I nt ensi t y and Hy dr aul i c Di amet er next t o t he Tur bul ence

Speci f i cat i on Met hod. Then ent er 1 for Tur bul ence I nt ensi t y and 0.2 for

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (5 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turblent Pipe Flow Step #2

Hy dr aul i c Di amet er . Click OK t o set t he velocit y.

The ( absolut e) pressure at t he out let is 1 at m. Since t he operat ing pressure is

set t o 1 at m, t he out let gauge pressure = out let absolut e pressure - operat ing

pressure = 0. Choose out l et under Zone. The Ty pe of t his boundary is

pr essur e- out l et . Click on Set .... The default value of t he Gauge Pr essur e is 0.

Click Cancel t o leave t he default s in place.

Not e: Backflow in t he Pressure Out let menu refers t o flow ent ering t hrough an

out let boundary. This is not likely t o happen in t his case. So we don' t have t o

set t he backflow paramet ers.

This complet es t he boundary condit ion specificat ion. Cl ose t he Boundary

Condit ions menu.

Go t o St ep 5: Solve!

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step4.htm (6 of 6)11/7/2005 6:49:33 PM

Fluent Tutorial - Turbulent Pipe Flow Step #3

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Sol ve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Step 5: Solve!

We' ll use second- order discret izat ion for t he moment um equat ion, as in t he

laminar pipe flow t ut orial, and also for t he t urbulence kinet ic energy equat ion

which is part of t he k- epsilon t urbulence model.

Ma in Men u > Solve > Con t r ols > Solu t ion . . .

Change Di scr et i zat i on for Moment um, Tur bul ence Ki net i c Ener gy and

Tur bul ence Di ssi pat i on Rat e ( scroll down t o see it ) equat ions t o Second

Or der Upw i nd.

Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step5.htm (1 of 5)11/7/2005 6:49:57 PM

Fluent Tutorial - Turbulent Pipe Flow Step #3

The order of discret izat ion t hat we j ust set refers t o t he convect ive t erms in

t he equat ions; t he discret izat ion of t he viscous t erms is always second- order

accurat e in FLUENT. Second- order discret izat ion generally yields bet t er

accuracy while first - order discret izat ion yields more robust convergence. I f

t he second- order scheme doesn' t converge, you can t ry st art ing t he it erat ions

wit h t he first - order scheme and swit ching t o t he second- order scheme aft er

some it erat ions.

Set I ni t i al Guess

We' ll use an init ial guess t hat is const ant over t he ent ire flow domain and equal

t o t he values at t he inlet :

Ma in Men u > Solve > I n it ia lize > I n it ia lize. . .

I n t he Solut ion I nit ializat ion menu t hat comes up, choose i nl et under Comput e

Fr om. The Ax i al Vel oci t y for all cells will be set t o 1 m/ s, t he Radi al Vel oci t y

t o 0 m/ s and t he Gauge Pr essur e t o 0 Pa. The Tur bul ence Ki net i c Ener gy

and Di ssi pat i on Rat e ( scroll down t o see it ) values are set from t he prescribed

values for t he Turbulence I nt ensit y and Hydraulic Diamet er at t he inlet .

Click I ni t . Close t he Solut ion I nit ializat ion window.

Set Conver gence Cr i t er i a

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step5.htm (2 of 5)11/7/2005 6:49:57 PM

Fluent Tutorial - Turbulent Pipe Flow Step #3

Recall t hat FLUENT report s a residual for each governing equat ion being solved.

The residual is a measure of how well t he current solut ion sat isfies t he discret e

form of each governing equat ion. We' ll it erat e t he solut ion unt il t he residual for

each equat ion falls below 1e- 6.

Ma in Men u > Solve > Mon it or s > Res id u a l. . .

Not ice t hat Conver gence Cr i t er i on has t o be set for t he k and epsilon

equat ions in addit ion t o t he t hree equat ions in t he last t ut orial. Set t he

Conver gence Cr i t er i on t o be 1e-06 for all five equat ions being solved.

Select Pr i nt and Pl ot under Opt i ons. This will print as well plot t he residuals as

t hey are calculat ed which you will use t o monit or convergence.

Click OK.

This complet es t he problem specificat ion. Save your work:

Ma in Men u > File > Wr it e > Ca s e. . .

Type in pipe100x30.cas for Case Fi l e. Click OK. Check t hat t he file has been

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step5.htm (3 of 5)11/7/2005 6:49:57 PM

Fluent Tutorial - Turbulent Pipe Flow Step #3

creat ed in your working direct ory.

I t er at e Unt i l Conver gence

Solve for 100 it erat ions first .

Ma in Men u > Solve > I t er a t e. . .

I n t he I t erat e menu t hat comes up, change t he Number of I t er at i ons t o 100.

Click I t er at e.

You' ll find t hat not all residuals have fallen below 1e- 6 in 100 it erat ions. Solve for

200 more it erat ions. The solut ion converges in a t ot al of 229 it erat ions.

( Click pict ure for larger image)

We need a larger number of it erat ions for convergence t han in t he laminar

case since we have a finer mesh and are also solving addit ional equat ions

from t he t urbulence model.

Save t he solut ion t o a dat a file:

Ma in Men u > File > Wr it e > Da t a . . .

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step5.htm (4 of 5)11/7/2005 6:49:57 PM

Fluent Tutorial - Turbulent Pipe Flow Step #3

Ent er pipe100x30.dat for Dat a Fi l e and click OK. Check t hat t he file has been

creat ed in your working direct ory.

Go t o St ep 6: Analyze Result s

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step5.htm (5 of 5)11/7/2005 6:49:57 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Anal y ze Resul t s

7. Refine Mesh

Problem 1

Step 6: Analyze Results

y +

Turbulent flows are significant ly affect ed by t he presence of walls. The k- epsilon

t urbulence model is primarily valid away from walls and special t reat ment is

required t o make it valid near walls. The near- wall model is sensit ive t o t he grid

resolut ion which is assessed in t he wall unit y+ ( defined in sect ion 10. 9. 1 of t he

FLUENT user manual) . We' ll gloss over t he det ails for now and use t he following

rule of t humb: select t he near- wall resolut ion such t hat y+ > 30 or < 5 for t he

wall- adj acent cell. Look at sect ion 10. 9, Grid Considerat ions for Turbulent Flow

Simulat ions, for det ails.

First , we need t o set t he reference values needed t o calculat e y+ .

Ma in Men u > Rep or t > Refer en ce Va lu es . . .

Select i nl et under Comput e Fr om t o t ell FLUENT t o use values at t he pipe inlet

for t he reference values. Check t hat t he reference value for densit y is 1 kg/ m3,

velocit y is 1 m/ s, and coefficient of viscosit y is 2e- 5 kg/ m- s as given in t he

Problem Specificat ion. These reference values will be used t o non- dimensionalize

t he dist ance of t he cell cent er from t he wall t o obt ain t he corresponding y+

values. Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (1 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

Let ' s plot y+ values for wall- adj acent cells t o check how it compares wit h t he

recommendat ion ment ioned above.

Ma in Men u > Plot > XY Plot . . .

Make sure t hat Posi t i on on X Ax i s is set under Opt i ons, t hat 1 is t he value

next t o X, and 0 is t he value next t o Y and Z under Pl ot Di r ect i on. Recall t hat

t his t ells FLUENT t o plot t he x- coordinat e value on t he abscissa of t he graph. Pick

Tur bul ence... under Y Ax i s Funct i on and select Wal l Ypl us from t he drop

down list under t hat . Since we want t he y+ value for cells adj acent t o t he wall of

t he pipe, choose w al l under Sur f aces.

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (2 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

Click Pl ot .

( Click pict ure for larger image)

As we can see, t he wall y+ value is bet ween 1. 6 and 1. 9 ( ignoring t he

anamolous at t he inlet ) . Since t his is less t han 5, t he near- wall grid resolut ion is

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (3 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

accept able.

Sav e Pl ot

I n t he Solut ion XY Plot Window, check t he Wr i t e t o Fi l e box under Opt i ons. The

Pl ot but t on should have changed t o t he Wr i t e... but t on. Click on Wr i t e.... Ent er

yplus.xy as t he filename and click OK. Check t hat t his file has been creat ed in

your FLUENT working direct ory.

Cent er l i ne Vel oci t y

Under Y Ax i s Funct i on, pick Vel oci t y ... and t hen in t he box under t hat , pick

Ax i al Vel oci t y . Finally, select cent er l i ne under Sur f aces since we are plot t ing

t he axial velocit y along t he cent erline. De- select w al l under Sur f aces.

Click on Cur ves... in t he Solut ion XY Plot window. Select t he solid line opt ion

under Pat t er n as shown below. Change Wei ght t o 2. Select t he blank opt ion

under Sy mbol . Click Appl y and Cl ose.

s

Turn on grid lines: I n t he Solut ion XY Plot window, click on Ax es.... Turn on t he

grid by checking t he boxes Maj or Rul es and Mi nor Rul es under Opt i ons. Click

Appl y . Select Y under Ax i s and repeat . Click Appl y and Cl ose.

Uncheck Wr i t e t o Fi l e. Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (4 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

( Click pict ure for larger image)

We can see t hat t he fully developed region st art s around x= 5m wit h t he

cent erline velocit y becoming const ant at a value of 1. 195 m/ s. This is quit e a bit

lower t han t he value of 2 m/ s for t he laminar case. Can you explain t he

difference based on t he physical charact erist ics of laminar and t urbulent flows?

Save t he dat a for t his plot as vel.xy.

Coef f i ci ent of Sk i n Fr i ct i on

The definit ion of t he skin frict ion coefficient was discussed in t he laminar pipe

flow t ut orial. The required reference values of densit y and velocit y have already

been set when plot t ing y+ .

Go back t o t he Solut ion XY Plot Window. Under t he Y Ax i s Funct i on, pick Wal l

Fl ux es..., and t hen Sk i n Fr i ct i on Coef f i ci ent in t he box under t hat . Under

Sur f aces, we are plot t ing t he frict ion coefficient along t he w al l . Uncheck

cent er l i ne surface.

Uncheck Wr i t e t o Fi l e. Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (5 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

( Click pict ure for larger image)

We can see t hat t he fully- developed value is 0. 0085. Compare t his wit h what

you' d expect from t he Moody chart .

Save t he dat a for t his plot as cf.xy.

Vel oci t y Pr of i l e

We' ll plot t he axial velocit y at t he out let as a funct ion of t he dist ance from t he

cent er of t he pipe.

Change t he plot set t ings so t hat t he radial dist ance from t he axis is plot t ed as

t he ordinat e: I n t he Solut ion XY Plot window, uncheck Posi t i on on X Ax i s under

Opt i ons and choose Posi t i on on Y Ax i s inst ead. Under Pl ot Di r ect i on, change

X t o 0 and Y t o 1. For t he X Ax i s Funct i on i. e. t he abscissa, pick Vel oci t y ...

and Ax i al Vel oci t y under t hat .

Since we want t o plot t his at t he out let boundary, pick only out l et under

Sur f aces.

Uncheck Wr i t e t o Fi l e. Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (6 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #4

( Click pict ure for larger image)

The axial velocit y is maximum at t he cent erline and zero at t he wall t o sat isfy t he

no- slip boundary condit ion for viscous flow. Compare qualit at ively t he near- wall

velocit y gradient normal t o t he wall wit h t he laminar case. Which is larger? From

t his, what can you say about t he relat ive st regt hs of near- wall mixing in t he

laminar and t urbulent cases?

Save t his plot as profile.xy.

Go t o St ep 7: Refine Mesh

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step6.htm (7 of 7)11/7/2005 6:50:43 PM

Fluent Tutorial - Turbulent Pipe Flow Step #5

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Ref i ne Mesh

Problem 1

Step 7: Refine Mesh

I n order t o assess t he numerical accuracy of t he result s obt ained, it is necessary

t o compare result s on different meshes. We' ll re- do t he calculat ion on a 100x60

mesh which has t wice t he number of nodes in t he radial direct ion as t he 100x30

mesh. You can download t he 100x60 mesh here.

File > Rea d > Ca s e. . .

Navigat e t o your working direct ory elect t he pipe100x60.msh file you have

creat ed. Click OK. Display t he grid. Check it s size.

Fi ner Mesh Anal y si s

Repeat st eps 4, 5, and 6 of t his t ut orial wit h t he finer mesh.

When you get t o st ep 6 of t he t ut orial, plot each of t he graphs as described.

However, for each of t he plot s, overlay t he corresponding result for t he coarser

mesh so t hat we may compare t hem. To do t his, aft er t he plot t ing t he finer mesh

result , in t he Solut ion XY Plot Window, click on Load Fi l e.... Navigat e t o your

working folder, click on t he appropriat e filename for t he previous result , eg. vel.

xy for cent erline velocit y, and click OK. Click Pl ot . You' ll see bot h result s plot t ed

in t he same t he graphics window.

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step7.htm (1 of 4)11/7/2005 6:50:56 PM

Fluent Tutorial - Turbulent Pipe Flow Step #5

( Click pict ure for larger image)

I n t he cent erline velocit y plot above, t he whit e line represent s t he cent erline

velocit y of t he finer mesh, while t he red line represent s t he velocit y of t he

coarser mesh from before. As we can see, t here isn' t t oo much of a difference

bet ween t he t wo plot s. Save t his plot as velt2.xy.

Now, let ' s t ake a look at t he coefficient of skin frict ion. This t ime, load t he cft.

xy file t o compare against t he plot . This is t he coefficient of skin frict ion plot :

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step7.htm (2 of 4)11/7/2005 6:50:56 PM

Fluent Tutorial - Turbulent Pipe Flow Step #5

( Click pict ure for larger image)

Once again, we can see t hat due t o t he fine degree of each mesh, t here isn' t

much difference bet ween t he t wo plot s. Save t his plot as cf2.xy. Now, st udy t he

velocit y of t he out let by plot t ing and comparing t o t he graph in outt.xy.

( Click pict ure for larger image)

Once again, t he finer mesh in t his case doesn' t offer much more precision t han

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step7.htm (3 of 4)11/7/2005 6:50:56 PM

Fluent Tutorial - Turbulent Pipe Flow Step #5

t he coarser mesh. Save t his plot as outt2.xy. Now let ' s t ake a look at t he YPlus

plot .

( Click pict ure for larger image)

As we can see, t here is a significant increase in t he accuracy of t he plot from t he

finer mesh. Save t his plot as yplus2.xy.

You may want t o experiment wit h meshes of ot her granularit ies and compare

t heir plot s wit h t he plot s saved from t he 100x30 and 100x60 meshes.

I n Problem 1, we will be looking at t he effect of coarse meshes wit h uniform

granularit y.

Go t o Problem 1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/step7.htm (4 of 4)11/7/2005 6:50:56 PM

Fluent Tutorial - Simple Pipe Flow

Tur bul ent Pi pe Fl ow

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Pr obl em 1

Problem 1

Pr obl em

Use FLUENT t o resolve t he developing flow in a pipe ( same configurat ion as was

done in t he t ut orial) for a pipe Reynolds number of 10, 000 on t he following

meshes: 100x5, 100x20 wit h uniform spacing in t he radial direct ion. Plot t he skin

frict ion c

f

as a funct ion of axial locat ion for each grid. Compare t he exit value

wit h t he expect ed value for fully developed flow ( e. g. , see Whit e pgs. 345- 346) .

Recall t hat a key quest ion for t he int egrit y of t he mesh is t he non- dimensional

value of t he first nodal point :

This should be eit her less t han 4 ( so t hat you resolve down int o t he viscous

sublayer) or great er t han 30 ( where wall funct ions can accurat ely compensat e

for t he poorly resolved viscous sublayer) . I nt ermediat e values can lead t o

great er errors. Calculat e t he value of y

1

+

for each mesh; use t hat t o help explain

( briefly) t he t rends in t he agreement t hat you observe.

Hi nt s

I f you no longer have t he 100x5 or 100x20 mesh, you can download t hem here:

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/ps1.htm (1 of 2)11/7/2005 6:51:03 PM

Fluent Tutorial - Simple Pipe Flow

pipe100x5. msh, pipe100x20. msh

Back t o Problem Specificat ion

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/pipe2/ps1.htm (2 of 2)11/7/2005 6:51:03 PM

Compressible Flow in a Nozzle - Step #1

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Cr eat e Geomet r y i n GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 1: Create Geometry in GAMBIT

Since t he nozzle has a circular cross- sect ion, it ' s reasonable t o assume t hat t he

flow is axisymmet ric. So t he geomet ry t o be creat ed is t wo- dimensional.

St ar t GAMBI T

Creat e a new direct ory called nozzle and st art GAMBI T from t hat direct ory by

t yping gambit -id nozzle at t he command prompt .

Under Ma in Men u , select Solver > FLUENT 5/ 6 since t he mesh t o be creat ed is

t o be used in FLUENT 6. 0.

Cr eat e Ax i s Edge

We' ll creat e t he bot t om edge corresponding t o t he nozzle axis by creat ing t he

vert ices A and B shown in t he above figure and j oining t hem by a st raight line.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ver t ex Com m a n d

Bu t t on > Cr ea t e Ver t ex

Creat e t he following t wo vert ices:

Vert ex 1: ( - 0. 5, 0, 0)

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (1 of 6)11/7/2005 6:51:18 PM

Compressible Flow in a Nozzle - Step #1

Vert ex 2: ( 0. 5, 0, 0)

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Cr ea t e Ed ge

Select vert ex 1 by holding down t he Shi f t but t on and clicking on it . Next , select

vert ex 2. Click Appl y in t he Creat e St raight Edge window.

Cr eat e Wal l Edge

We' ll next creat e t he bot t om edge corresponding t o t he nozzle wall. This edge is

curved. Since

A= pi r

2

where r( x) is t he radius of t he cross- sect ion at x and

A = 0. 1 + x

2

for t he given nozzle geomet ry, we get

r( x) = [ ( 0. 1 + x

2

) / pi]

0. 5

; - 0. 5 < x < 0. 5

This is t he equat ion of t he curved wall. Life would have been easier if GAMBI T

allowed for t his equat ion t o be ent ered direct ly t o creat e t he curved edge.

I nst ead, one has t o creat e a file cont aining t he coordinat es of a series of point s

along t he curved line and read in t he file. The more number of point s used along

t he curved edge, t he smoot her t he result ant edge.

The file vert . dat cont ains t he point definit ions for t he nozzle wall. Take a look at

t his file. The first line is

21 1

which says t hat t here are 21 point s along t he edge and we are defining only 1

edge. This is followed by x, r and z coordinat es for each point along t he edge. The

r- value for each x was generat ed from t he above equat ion for r( x) . The z-

coordinat e is 0 for all point s since we have a 2D geomet ry.

Ri ght - cl i ck on vert . dat and select Save As... t o download t he file t o your

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (2 of 6)11/7/2005 6:51:18 PM

Compressible Flow in a Nozzle - Step #1

working direct ory.

Ma in Men u > File > I n p u t > I CEM I n p u t . . .

Next t o Fi l e Name: , ent er t he pat h t o t he vert.dat file t hat you downloaded or

browse t o it by clicking on t he Br ow se but t on.

Then, check t he Ver t i cesand Edges boxes under Geomet r y t o Cr eat e as we

want t o creat e t he vert ices as well as t he curved edge.

Click Accept .

This should creat e t he curved edge. Here it is in relat ion t o t he vert ices we

creat ed above:

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (3 of 6)11/7/2005 6:51:18 PM

Compressible Flow in a Nozzle - Step #1

( Click pict ure for larger image)

Cr eat e I nl et and Out l et Edges

Creat e t he vert ical edge for t he inlet :

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Cr ea t e Ed ge

Shi f t - cl i ck on vert ex 1 and t hen t he vert ex above it t o creat e t he inlet edge.

Similarly, creat e t he vert ical edge for t he out let .

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (4 of 6)11/7/2005 6:51:18 PM

Compressible Flow in a Nozzle - Step #1

( Click pict ure for larger image)

Cr eat e Face

Form a face out of t he area enclosed by t he four edges:

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on

> For m Fa ce

Recall t hat we have t o shi f t - cl i ck on each of t he edges enclosing t he face and

t hen click Appl y t o creat e t he face.

Sav e Your Wor k

Ma in Men u > File > Sa ve

This will creat e t he nozzle. dbs file in your working direct ory. Check t hat it has

been creat ed so t hat you will able t o resume from here if necessary.

Go t o St ep 2: Mesh Geomet ry in GAMBI T

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (5 of 6)11/7/2005 6:51:18 PM

Compressible Flow in a Nozzle - Step #1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step1.htm (6 of 6)11/7/2005 6:51:18 PM

Fluent Tutorial - Compressible Flow in a Nozzle Step #2

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet r y i n GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 2: Mesh Geometry in GAMBIT

Now t hat we have t he basic geomet ry of t he nozzle creat ed, we need t o mesh it .

We would like t o creat e a 50x20 grid for t his geomet ry.

Mesh Edges

As in t he previous t ut orials, we will first st art by meshing t he edges.

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on >

Mes h Ed ges

Like t he Laminar Pipe Flow Tut orial, we are going t o use even spacing bet ween

each of t he mesh point s. We won' t be using t he Grading t his t ime, so deselect

t he box next t o Gr adi ng t hat says Appl y .

Then, change I nt er val Count t o 20 for t he side edges and I nt er val Count t o

50 for t he t op and bot t om edges.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step2.htm (1 of 3)11/7/2005 6:51:31 PM

Fluent Tutorial - Compressible Flow in a Nozzle Step #2

( Click pict ure for larger image)

Mesh Face

Now t hat we have t he edges meshed, we need t o mesh t he face.

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Mes h Fa ces

As before, select t he face and click t he Appl y but t on.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step2.htm (2 of 3)11/7/2005 6:51:31 PM

Fluent Tutorial - Compressible Flow in a Nozzle Step #2

( Click pict ure for large image)

Sav e Your Wor k

Ma in Men u > File > Sa ve

Go t o St ep 3: Specify Boundary Types in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step2.htm (3 of 3)11/7/2005 6:51:31 PM

Compressible Flow in a Nozzle - Step #3

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Speci f y Boundar y Ty pes i n GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 3: Specify Boundary Types in GAMBIT

Speci f y Boundar y Ty pes

Now t hat we have t he mesh, we would like t o specify t he boundary condit ions

here in GAMBI T.

Op er a t ion Toolp a d > Zon es Com m a n d Bu t t on > Sp ecify Bou n d a r y Typ es

Com m a n d Bu t t on

This will bring up t he Specify Boundary Types window on t he Operat ion Panel.

We will first specify t hat t he left edge is t he inlet . Under Ent i t y : , pick Edges so

t hat GAMBI T knows we want t o pick an edge ( face is default ) .

Now select t he left edge by Shi f t - cl i ck i ng on it . The select ed edge should

appear in t he yellow box next t o t he Edges box you j ust worked wit h as well as

t he Label / Ty pe list right under t he Edges box.

Next t o Name: , ent er inlet.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step3.htm (1 of 3)11/7/2005 6:51:51 PM

Compressible Flow in a Nozzle - Step #3

For Ty pe: , select VELOCI TY_I NLET.

Click Appl y . You should see t he new ent ry appear under Name/ Ty pe box near

t he t op of t he window.

Creat e boundary t ypes for each of t he edges as specified in t he chart below:

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step3.htm (2 of 3)11/7/2005 6:51:51 PM

Compressible Flow in a Nozzle - Step #3

Edge

Position

Name Type

Left inlet PRESSURE_INLET

Right outlet PRESSURE_OUTLET

Top wall WALL

Bottom centerline AXIS

You should have t he following edges in t he Name/ Ty pe list when finished:

Sav e and Ex por t

Ma in Men u > File > Sa ve

Ma in Men u > File > Exp or t > Mes h . . .

Type in nozzle.msh for t he Fi l e Name: . Select Ex por t 2d Mesh since t his is a 2

dimensional mesh. Click Accept .

Check nozzle.msh has been creat ed in your working direct ory.

Go t o St ep 4: Set Up Problem in FLUENT

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step3.htm (3 of 3)11/7/2005 6:51:51 PM

Compressible Flow in a Nozzle - Step #4

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Pr obl em i n FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 4: Set Up Problem in FLUENT

Launch FLUENT

St a r t > Pr ogr a m s > Flu en t I n c > FLUENT 6. 0

Select 2ddp from t he list of opt ions and click Run.

I mpor t Fi l e

Ma in Men u > File > Rea d > Ca s e. . .

Navigat e t o your working direct ory and select t he nozzle.msh file. Click OK.

The following should appear in t he FLUENT window:

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (1 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #4

Check t hat t he displayed informat ion is consist ent wit h our expect at ions of t he

nozzle grid.

Anal y ze Gr i d

Gr id > I n fo > Size

How many cells and nodes does t he grid have?

Dis p la y > Gr id

How many nodes are t here in t he radial direct ion? Are t he nodes clust ered

t owards t he wall? Why?

Def i ne Pr oper t i es

Defin e > Mod els > Solver . . .

Under t he Sol ver box, select Coupl ed. Under Space, choose Ax i sy mmet r i c.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (2 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #4

Click OK.

Defin e > Mod els > Vis cou s

Select I nvi sci d under Model .

Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (3 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #4

Defin e > Mod els > En er gy

The energy equat ion needs t o be t urned on since t his is a compressible flow

where t he energy equat ion is coupled t o t he cont inuit y and moment um

equat ions.

Make sure t here is a check box next t o Ener gy Equat i on and click OK.

Defin e > Ma t er ia ls

Select ai r under Fl ui d mat er i al s. Under Pr oper t i es, choose I deal Gas next t o

Densi t y . You should see t he window expand. This means FLUENT uses t he ideal

gas equat ion of st at e t o relat e densit y t o t he st at ic pressure and t emperat ure.

Click Change/ Cr eat e.

Defin e > Op er a t in g Con d it ion s

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (4 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #4

We' ll work in t erms of absolut e rat her t han gauge pressures in t his example. So

set Oper at i ng Pr essur e in t he Pr essur e box t o 0.

Click OK.

I t is import ant t hat you set t he operat ing pressure correct ly in compressible flow

calculat ions since FLUENT uses it t o comput e absolut e pressure t o use in t he

ideal gas law.

Defin e > Bou n d a r y Con d it ion s

Set boundary condit ions for t he following surfaces: axis, default - int erior, fluid,

inlet , out let , wall.

Select i nl et under Sur f ace and pick pr essur e- i nl et under Ty pe as it s boundary

condit ion. Click Set .... The Pressure I nlet window should come up.

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (5 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #4

Set t he t ot al ( i. e. st agnat ion) pressure ( not ed as Gauage Tot al Pr essur e in

FLUENT) and t emperat ure at t he inlet . For a subsonic inlet , Super soni c/ I ni t i al

Gauge Pr essur e is t he init ial guess value for t he st at ic pressure. Calculat e t his

init ial guess value from t he 1D solut ion. Aft er you have ent ered t he values, click

OK t o close t he window.

Using t he same st eps as above, pick pr essur e- out l et as t he boundary condit ion

for t he out l et surface. Then, when t he Pressure Out let window comes up, set t he

pressure and t emperat ure as above. Click OK.

Go t o St ep 5: Solve!

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step4.htm (6 of 6)11/7/2005 6:52:36 PM

Compressible Flow in a Nozzle - Step #5

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Sol ve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 5: Solve!

Now we will set t he solve set t ings for t his problem and t hen it erat e t hrough and

act ually solve it .

Solve > Con t r ol > Solu t ion

Take a look at t he opt ions available. We want Second Or der Upw i nd for t he

Fl ow ( under t he Di scr et i zat i on box) .

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step5.htm (1 of 4)11/7/2005 6:52:52 PM

Compressible Flow in a Nozzle - Step #5

Make sure t hat is select ed and click OK.

Solve > I n it ia lize

As you may recall from t he previous t ut orials, t his is where we set t he init ial

guess values ( t he base case) for t he it erat ive solut ion. Once again, we' ll set

t hese values t o be t he ones at t he inlet . Select i nl et under Comput e Fr om.

Click I ni t .

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step5.htm (2 of 4)11/7/2005 6:52:52 PM

Compressible Flow in a Nozzle - Step #5

Solve > Mon it or s > Res id u a l

Now we will set t he residual values ( t he crit eria for a good enough solut ion) .

Once again, we' ll set t his value t o 1e- 06.

Click OK.

Solve > I t er a t e

What does t he convergence plot look like?

How many it erat ions does it t ake t o converge?

Save case and dat a aft er you have obt ained a converged solut ion.

Go t o St ep 6: Analyze Result s

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step5.htm (3 of 4)11/7/2005 6:52:52 PM

Compressible Flow in a Nozzle - Step #5

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step5.htm (4 of 4)11/7/2005 6:52:52 PM

Compressible Flow in a Nozzle - Step #6

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Anal y ze Resul t s

7. Refine Mesh

Problem 1

Problem 2

Step 6: Analyze Results

Mach Number Pl ot

As in t he previous t ut orials, we are going t o plot t he velocit y along t he

cent erline. However, t his t ime, we are going t o use t he dimensionless Mach

quant it y.

Plot > XY Plot

We are going plot t he variat ion of t he Mach number in t he axial direct ion at t he

axis and wall. I n addit ion, we will plot t he corresponding variat ion from 1D

t heory. You can download t he file here: mach_1D. xy.

Do everyt hing as we would do for plot t ing t he cent erline velocit y. However,

inst ead of select ing Ax i al Vel oci t y as t he Y Ax i s Funct i on, select Mach

Number .

Also, since we are going t o plot t his number at bot h t he wall and axis, select

cent er l i ne and w al l under Sur f aces.

Then, load t he mach_1D. xy by clicking on Load Fi l e....

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step6.htm (1 of 5)11/7/2005 6:53:25 PM

Compressible Flow in a Nozzle - Step #6

Click Pl ot .

( Click pict ure for large image)

How does t he FLUENT solut ion compare wit h t he 1D solut ion?

I s t he comparison bet t er at t he wall or at t he axis? Can you explain t his?

Save t his plot as machplot.xy by checking Wr i t e t o Fi l e and clicking Wr i t e....

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step6.htm (2 of 5)11/7/2005 6:53:25 PM

Compressible Flow in a Nozzle - Step #6

Pr essur e Cont our Pl ot

Somet imes, it is very useful t o see how t he pressure and t emperat ure changes

t hroughout t he obj ect . This can be done via cont our plot s.

Dis p la y > Con t ou r s . . .

First , we are going t o plot t he pressure cont ours of t he nozzle. Therefore, make

sure t hat under Cont our s Of , Pr essur e... and St at i c Pr essur e is select ed.

We want t his at a fine enough granularit y so t hat we can see t he pressure

changes clearly. Under Level s, change t he default 20 t o 40. This increases t he

number of lines in t he cont our plot so t hat we can get a more accurat e result .

Click Di spl ay .

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step6.htm (3 of 5)11/7/2005 6:53:25 PM

Compressible Flow in a Nozzle - Step #6

( Click pict ure for large image)

Not ice t hat t he pressure on t he fluid get s smaller as it flows t o t he right , as is

consist ent wit h fluid going t hrough a nozzle.

Temper at ur e Cont our Pl ot

Now we will plot t he t emperat ure cont ours and see how t he t emperat ure varies

t hroughout t he nozzle.

Back in t he Cont ours window, under Cont our s Of , select Temper at ur e... and

St at i c Temper at ur e.

Click Di spl ay .

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step6.htm (4 of 5)11/7/2005 6:53:25 PM

Compressible Flow in a Nozzle - Step #6

( Click pict ure for large image)

As we can see, t he t emperat ure decreases t owards t he right side of t he nozzle,

indicat ing a change of int ernal energy t o kinet ic energy as t he fluid speeds up.

Go t o St ep 7: Refine Mesh

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step6.htm (5 of 5)11/7/2005 6:53:25 PM

Compressible Flow in a Nozzle - Step #7

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Ref i ne Mesh

Problem 1

Problem 2

Step 7: Refine Mesh

Solve t he nozzle flow for t he same condit ions as used in class on a 80x30 grid.

Recall t hat t he st at ic pressure p at t he exit is 3, 738. 9 Pa. The grid for t his

calculat ion can be downloaded here. You may also download it from here.

( a) Plot t he variat ion of Mach number at t he axis and t he wall as a funct ion of

t he axial dist ance x. Also, plot t he corresponding result s obt ained on t he 50x20

grid used in class and from t he quasi- 1D assumpt ion. Recall t hat t he quasi- 1D

result for t he Mach number variat ion was given t o you in t he M_1D.xy file. Not e

all five curves should be plot t ed on t he same graph so t hat you can compare

t hem. You can make t he plot s in FLUENT, MATLAB or EXCEL.

( b) Plot t he variat ion of st at ic pressure at t he axis and t he wall as a funct ion of

t he axial dist ance x. Also, plot t he corresponding result s obt ained on t he 50x20

grid used in class and from t he quasi- 1D assumpt ion. Calculat e t he st at ic

pressure variat ion for t he quasi- 1D case from t he Mach number variat ion given

in M_1D.xy.

( c) Plot t he variat ion of st at ic t emperat ure at t he axis and t he wall as a funct ion

of t he axial dist ance x. Also, plot t he corresponding result s obt ained on t he

50x20 grid used in class and from t he quasi- 1D assumpt ion. Calculat e t he st at ic

t emperat ure variat ion for t he quasi- 1D case from t he Mach number variat ion

given in M_1D.xy.

Comment very briefly on t he grid dependence of your result s and t he comparison

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step7.htm (1 of 2)11/7/2005 6:53:49 PM

Compressible Flow in a Nozzle - Step #7

wit h t he quasi- 1D result s.

Go t o Problem 1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/step7.htm (2 of 2)11/7/2005 6:53:49 PM

Compressible Flow in a Nozzle - Problem #1

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Grid

Pr obl em 1

Problem 2

Problem 1

Consider t he nozzle flow problem solved using FLUENT in t he t ut orial. Recall t hat

t he nozzle has a circular cross- sect ional area, A, t hat varies wit h axial dist ance

from t he t hroat , x, according t o t he formula:

A = 0. 1 + x

2

where A is in square met ers and x is in met ers. The st agnat ion pressure p

o

and

st agnat ion t emperat ure T

o

at t he inlet are 101, 325 Pa and 300 K, respect ively.

Using t he quasi- 1D flow assumpt ion, det ermine t he st at ic pressure at t he nozzle

inlet and out let for t he following condit ions:

( a) Sonic flow at t he t hroat , and supersonic, isent ropic flow in t he diverging

sect ion.

( b) Sonic flow at t he t hroat , and subsonic, isent ropic flow in t he diverging sect ion.

( c) Sonic flow at t he t hroat and normal shock at t he exit .

Go t o Problem 2

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/ps1.htm (1 of 2)11/7/2005 6:54:14 PM

Compressible Flow in a Nozzle - Problem #1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/ps1.htm (2 of 2)11/7/2005 6:54:14 PM

Compressible Flow in a Nozzle - Problem 3

Compr essi bl e Fl ow i n a Nozzl e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Grid

Problem 1

Pr obl em 2

Problem 2

Change t he exit pressure t o 40, 000 Pa while keeping all t he ot her boundary

condit ions t he same. What flow regime do you expect for t his exit pressure

based on t he quasi- 1D result s in problem 1? Re- run t he FLUENT calculat ion wit h

t his exit pressure on t he 50x20 grid.

( a) Plot cont ours of t he Mach number and st at ic pressure for t his case. I s t he

flow regime as predict ed by quasi- 1D t heory? Explain briefly t he possible causes

for any similarit ies or disparit ies.

( b) Plot t he st at ic and st agnat ion pressures at t he axis as a funct ion of t he axial

dist ance. Also, plot t he corresponding values from t he case where t he exit

pressure is 3, 738. 9 Pa. ( These four curves should be on t he same graph. )

Explain briefly t he salient feat ures of t his plot .

( c) Plot t he st at ic and st agnat ion t emperat ures at t he axis as a funct ion of t he

axial dist ance. Again provide a brief explanat ion for t he salient feat ures.

Back t o Problem Specificat ion

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/ps2.htm (1 of 2)11/7/2005 6:54:37 PM

Compressible Flow in a Nozzle - Problem 3

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/nozzle/ps2.htm (2 of 2)11/7/2005 6:54:37 PM

Flow over an Airfoil - Step #1

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Cr eat e Geomet r y i n GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 1: Create Geometry in GAMBIT

I f you wish t o skip t he st eps for grid creat ion, you can download t he mesh file

here ( r i ght - cl i ck and select Save As...) and go t o St ep 4.

This t ut orial leads you t hrough t he st eps for generat ing a mesh in GAMBI T for an

airfoil geomet ry. This mesh can t hen be read int o FLUENT for fluid flow

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (1 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

simulat ion.

I n an ext ernal flow such as t hat over an airfoil, we have t o define a farfield

boundary and mesh t he region bet ween t he airfoil geomet ry and t he farfield

boundary. I t is a good idea t o place t he farfield boundary well away from t he

airfoil since we' ll use t he ambient condit ions t o define t he boundary condit ions at

t he farfield. The fart her we are from t he airfoil, t he less effect it has on t he flow

and so more accurat e is t he farfield boundary condit ion.

The farfield boundary we' ll use is t he line ABCDEFA in t he figure above. c is t he

chord lengt h.

St ar t GAMBI T

Creat e a new direct ory called airfoil and st art GAMBI T from t hat direct ory by

t yping gambit -id airfoil at t he command prompt .

Under Ma in Men u , select Solver > FLUENT 5/ 6 since t he mesh t o be creat ed is

t o be used in FLUENT 6. 0.

I mpor t Edge

To specify t he airfoil geomet ry, we' ll import a file cont aining a list of vert ices

along t he surface and have GAMBI T j oin t hese vert ices t o creat e t wo edges,

corresponding t o t he upper and lower surfaces of t he airfoil. We' ll t hen split t hese

edges int o 4 dist inct edges t o help us cont rol t he mesh size at t he surface.

The file cont aining t he vert ices for t he airfoil can be downloaded here: vert ices.

dat ( right click and select Save As...)

Let ' s t ake a look at t he vert ices. dat file:

The first line of t he file represent s t he number of point s on each edge ( 61) and

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (2 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

t he number of edges ( 2) . The first 61 set of vert ices are connect ed t o form t he

edge corresponding t o t he upper surface; t he next 61 are connect ed t o form t he

edge for t he lower surface.

The chord lengt h c for t he geomet ry in vert ices. dat file is 1, so x varies bet ween

0 and 1. I f you are using a different airfoil geomet ry specificat ion file, not e t he

range of x values in t he file and det ermine t he chord lengt h c. You' ll need t his

lat er on.

Ma in Men u > File > I m p or t > I CEM I n p u t . . .

For Fi l e Name, browse and select t he vert ices. dat file. Select bot h Ver t i ces and

Edges under Geomet r y t o Cr eat e: since t hese are t he geomet ric ent it ies we

need t o creat e. Deselect Face. Click Accept .

( Click pict ure for larger image)

Spl i t Edges

Next , we will split t he t op and bot t om edges int o t wo edges each as shown in t he

figure below.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (3 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

We need t o do t his because a non- uniform grid spacing will be used for x< 0. 3c

and a uniform grid spacing for x> 0. 3c. To split t he t op edge int o HI and I G, select

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Sp lit / Mer ge Ed ge

Make sure Poi nt is select ed next t o Spl i t Wi t h in t he Split Edge window.

Select t he t op edge of t he airfoil by Shi f t - cl i ck i ng on it . You should see

somet hing similar t o t he pict ure below:

( Click pict ure for larger image)

We' ll use t he point at x= 0. 3c on t he upper surface t o split t his edge int o HI and

I G. To do t his, ent er 0.3 for x : under Gl obal . I f your c is not equal t o one, ent er

t he value of 0. 3* c inst ead of j ust 0. 3. For inst ance, if c= 4, ent er 1.2. From here

on, whenever you' re asked t o ent er ( some fact or) * c, calculat e t he appropriat e

value for your c and ent er it .

You should see t hat t he whit e circle has moved t o t he correct locat ion on t he

edge.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (4 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

( Click pict ure for larger image)

Click Appl y . You will see a message saying ` ` Edge edge. 1 was split , and edge

edge. 3 creat ed' ' in t he Transcript window.

( Click pict ure for larger image)

Not e t he yellow marker in place of t he whit e circle, indicat ing t he original edge

has been split int o t wo edges wit h t he yellow marker as it s dividing point .

Repeat t his procedure for t he lower surface t o split it int o HJ and JG. Use t he

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (5 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

point at x= 0. 3c on t he lower surface t o split t his edge.

Cr eat e Far f i el d Boundar y

Next we' ll creat e t he farfield boundary by creat ing vert ices and j oining t hem

appropriat ely t o form edges.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ver t ex Com m a n d

Bu t t on > Cr ea t e Ver t ex

Creat e t he following vert ices by ent ering t he coordinat es under Gl obal and t he

label under Label :

Label x-coordinate y-coordinate z-coordinate

A c 12.5c 0

B 21c 12.5c 0

C 21c 0 0

D 21c -12.5c 0

E c -12.5c 0

F -11.5c 0 0

G c 0 0

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (6 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

( Click pict ure for larger image)

Click t he FI T TO WI NDOW but t on t o scale t he display so t hat you can see all t he

vert ices.

As you creat e t he edges for t he farfield boundary, keep t he pict ure of t he farfield

nomenclat ure given at t he t op of t his st ep handy.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Cr ea t e Ed ge

Creat e t he edge AB by select ing t he vert ex A followed by vert ex B. Ent er AB for

Label . Click Appl y . GAMBI T will creat e t he edge. You will see a message saying

somet hing like "Creat ed edge: AB' ' in t he Transcript window.

Similarly, creat e t he edges BC, CD, DE, EG, GA and CG. Not e t hat you might

have t o zoom in on t he airfoil t o select vert ex G correct ly.

Next we' ll creat e t he circular arc AF. Ri ght - cl i ck on t he Cr eat e Edge but t on and

select Ar c.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (7 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

I n t he Creat e Real Circular Arc menu, t he box next t o Cent er will be yellow. That

means t hat t he vert ex you select will be t aken as t he cent er of t he arc. Select

vert ex G and click Appl y .

Now t he box next t o End Poi nt s will be highlight ed in yellow. This means t hat

you can now select t he t wo vert ices t hat form t he end point s of t he arc. Select

vert ex A and t hen vert ex F. Ent er AF under Label . Click Appl y .

I f you did t his right , t he arc AF will be creat ed. I f you look in t he t ranscript

window, you' ll see a message saying t hat an edge has been creat ed.

Similarly, creat e an edge corresponding t o arc EF.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (8 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

( Click pict ure for larger image)

Cr eat e Faces

The edges can be j oined t oget her t o form faces ( which are planar surfaces in

2D) . We' ll creat e t hree faces: ABCGA, EDCGE and GAFEG+ airfoil surface. Then

we' ll mesh each face.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on

> For m Fa ce

This brings up t he Creat e Face From Wireframe menu. Recall t hat we had

select ed vert ices in order t o creat e edges. Similarly, we will select edges in order

t o form a face.

To creat e t he face ABCGA, select t he edges AB, BC, CG, and GA and click Appl y .

GAMBI T will t ell you t hat it has "Creat ed face: face. 1' ' in t he t ranscript window.

Similarly, creat e t he face EDCGE.

To creat e t he face consist ing of GAFEG+ airfoil surface, select t he edges in t he

following order: AG, AF, EF, EG, and JG, HJ, HI and I G ( around t he airfoil in t he

clockwise direct ion) . Click Appl y .

Go t o St ep 2: Mesh Geomet ry in GAMBI T

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (9 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step1.htm (10 of 10)11/7/2005 6:54:58 PM

Flow over an Airfoil - Step #2

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet r y i n GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 2: Mesh Geometry in GAMBIT

Mesh Faces

We' ll mesh each of t he 3 faces separat ely t o get our final mesh. Before we mesh

a face, we need t o define t he point dist ribut ion for each of t he edges t hat form

t he face i. e. we first have t o mesh t he edges. We' ll select t he mesh st ret ching

paramet ers and number of divisions for each edge based on t hree crit eria:

1. We'd like to cluster points near the airfoil since this is where the flow is modified the most; the

mesh resolution as we approach the farfield boundaries can become progressively coarser since

the flow gradients approach zero.

2. Close to the surface, we need the most resolution near the leading and trailing edges since these

are critical areas with the steepest gradients.

3. We want transitions in mesh size to be smooth; large, discontinuous changes in the mesh size

significantly decrease the numerical accuracy.

The edge mesh paramet ers we' ll use for cont rolling t he st ret ching are successive

rat io, first lengt h and last lengt h. Each edge has a direct ion as indicat ed by t he

arrow in t he graphics window. The successive rat io R is t he rat io of t he lengt h of

any t wo successive divisions in t he arrow direct ion as shown below. Go t o t he

index of t he GAMBI T User Guide and look under Edge> Meshi ng for t his figure

and accompanying explanat ion. This help page also explains what t he first and

last lengt hs are; make sure you underst and what t hey are.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (1 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #2

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on >

Mes h Ed ges

Select t he edge GA. The edge will change color and an arrow and several circles

will appear on t he edge. This indicat es t hat you are ready t o mesh t his edge.

Make sure t he arrow is point ing upwards. You can reverse t he direct ion of t he

edge by clicking on t he Rever se but t on in t he Mesh Edges menu. Ent er a rat io of

1. 15. This means t hat each successive mesh division will be 1. 15 t imes bigger in

t he direct ion of t he arrow. Select I nt er val Count under Spaci ng. Ent er 45 for

I nt er val Count . Click Appl y . GAMBI T will creat e 45 int ervals on t his edge wit h

a successive rat io of 1. 15.

For edges AB and CG, we' ll set t he First Lengt h ( i. e. t he lengt h of t he division at

t he st art of t he edge) rat her t han t he Successive Rat io. Repeat t he same st eps

for edges BC, AB and CG wit h t he following specificat ions:

Edges Arrow Direction Successive Ratio Interval Count

GA and BC Upwards 1.15 45

Edges Arrow Direction First Length Interval Count

AB and CG Left to Right 0.02c 60

Not e t hat lat er we' ll select t he lengt h at t he t railing edge t o be 0. 02c so t hat t he

mesh lengt h is cont inuous bet ween I G and CG, and HG and CG.

Now t hat t he appropriat e edge meshes have been specified, mesh t he face

ABCGA:

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (2 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #2

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Mes h Fa ces

Select t he face ABCGA. The face will change

color. You can use t he default s of Quad ( i. e. quadrilat erals) and Map. Click

Appl y .

The meshed face should look as follows:

( Click pict ure for larger image)

Next mesh face EDCGE in a similar fashion. The following t able shows t he

paramet ers t o use for t he different edges:

Edges Arrow Direction Successive Ratio Interval Count

EG and CD Downwards 1.15 45

Edges Arrow Direction First Length Interval Count

DE Left to Right 0.02c 60

The result ant mesh should be symmet ric about CG as shown in t he figure below.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (3 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #2

( Click pict ure for larger image)

Finally, let ' s mesh t he face consist ing of GAFEG and t he airfoil surface. For edges

HI and HJ on t he front part of t he airfoil surface, use t he following paramet ers t o

creat e edge meshes:

Edges Arrow Direction Last Length Interval Count

HI From H to I 0.02c 40

HJ From H to J 0.02c 40

For edges I G and JG, we' ll set t he divisions t o be uniform and equal t o 0. 02c.

Use I nt erval Size rat her t han I nt erval Count and creat e t he edge meshes:

Edges Arrow Direction Successive Ratio Interval Size

IG and JG Left to Right 1 0.02c

For edge AF, t he number of divisions needs t o be equal t o t he number of

divisions on t he line opposit e t o it i. e. t he upper surface of t he airfoil ( t his is a

subt le point ; chew over it ) . To det ermine t he number of divisions t hat GAMBI T

has creat ed on edge I G, select

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

>Su m m a r ize Ed ge Mes h

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (4 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #2

Select edge I G and t hen El ement s under Component and click Appl y . This will

give t he t ot al number of nodes ( i. e. point s) and element s ( i. e. divisions) on t he

edge in t he Transcript window. The number of divisions on edge I G is 35. ( I f you

are using a different geomet ry, t his number will be different ; I ' ll refer t o it as

N

I G

) . So t he I nt erval Count for edge AF is N

HI

+ N

I G

= 40+ 35= 75.

Similarly, det ermine t he number of divisions on edge JG. This also comes out as

35 for t he current geomet ry. So t he I nt erval Count for edge EF also is 75.

Creat e t he mesh for edges AF and EF wit h t he following paramet ers:

Edges Arrow Direction First Length Interval Count

AF From A to F 0.02c

40+N

IG

EF From E to F 0.02c

40+N

JG

Mesh t he face. The result ant mesh is shown below.

( Click pict ure for larger image)

Go t o St ep 3: Specify Boundary Types in GAMBI T

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (5 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #2

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step2.htm (6 of 6)11/7/2005 6:55:22 PM

Flow over an Airfoil - Step #3

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Speci f y Boundar y Ty pes i n GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 3: Specify Boundary Types in GAMBIT

We' ll label t he boundary AFE as farfield1, ABDE as farfield2 and t he airfoil surface

as airfoil. Recall t hat t hese will be t he names t hat show up under boundary zones

when t he mesh is read int o FLUENT.

Gr oup Edges

We' ll creat e groups of edges and t hen creat e boundary ent it ies from t hese

groups.

First , we will group AF and EF t oget her.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Gr ou p Com m a n d Bu t t on

> Cr ea t e Gr ou p

Select Edges and ent er farfield1 for Label , which is t he name of t he group.

Select t he edges AF and EF.

Not e t hat GAMBI T adds t he edge t o t he list as it is select ed in t he GUI .

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step3.htm (1 of 4)11/7/2005 6:55:59 PM

Flow over an Airfoil - Step #3

Click Appl y .

I n t he t ranscript window, you will see t he message “ Creat ed group: farfield1

group” .

Similarly, creat e t he ot her t wo farfield groups. You should have creat ed a t ot al of

t hree groups:

Group Name Edges in Group

farfield1 AF,EF

farfield2 AB,DE

farfield3 BC,CD

airfoil HI,IG,HJ,JG

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step3.htm (2 of 4)11/7/2005 6:55:59 PM

Flow over an Airfoil - Step #3

Def i ne Boundar y Ty pes

Now t hat we have grouped each of t he edges int o t he desired groups, we can

assign appropriat e boundary t ypes t o t hese groups.

Op er a t ion Toolp a d > Zon es Com m a n d Bu t t on > Sp ecify Bou n d a r y Typ es

Under Ent i t y , select Gr oups.

Select any edge belonging t o t he airfoil surface and t hat will select t he airfoil

group. Next t o Name: , ent er airfoil. Leave t he Ty pe as WALL.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step3.htm (3 of 4)11/7/2005 6:55:59 PM

Flow over an Airfoil - Step #3

Click Appl y .

I n t he Transcript Window, you will see a message saying "Creat ed Boundary

ent it y: airfoil".

Similarly, creat e boundary ent it ies corresponding t o farfield1, farfield2 and

farfield3 groups. Set t he Ty pe t o Pr essur e Far f i el d in each case.

Sav e Your Wor k

Ma in Men u > File > Sa ve

Ex por t Mesh

Ma in Men u > File > Exp or t > Mes h . . .

Save t he file as airfoil.msh.

Make sure t hat t he Ex por t 2d Mesh opt ion is select ed.

Check t o make sure t hat t he file is creat ed.

Go t o St ep 4: Set Up Problem in FLUENT

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step3.htm (4 of 4)11/7/2005 6:55:59 PM

Flow over an Airfoil - Step #4

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Pr obl em i n FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 4: Set Up Problem in FLUENT

Launch FLUENT

St a r t > Pr ogr a m s > Flu en t I n c > FLUENT 6. 0

Select 2ddp from t he list of opt ions and click Run.

I mpor t Fi l e

Ma in Men u > File > Rea d > Ca s e. . .

Navigat e t o your working direct ory and select t he airfoil.msh file. Click OK.

The following should appear in t he FLUENT window:

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (1 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #4

Check t hat t he displayed informat ion is consist ent wit h our expect at ions of t he

airfoil grid.

Anal y ze Gr i d

Gr id > I n fo > Size

How many cells and nodes does t he grid have?

Dis p la y > Gr id

Not e what t he surfaces farfield1, farfield2, et c. correspond t o by select ing and

plot t ing t hem in t urn.

Zoom int o t he airfoil.

Where are t he nodes clust ered? Why?

Def i ne Pr oper t i es

Defin e > Mod els > Solver . . .

Under t he Sol ver box, select Segr egat ed.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (2 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #4

Click OK.

Defin e > Mod els > Vis cou s

Select I nvi sci d under Model .

Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (3 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #4

Defin e > Mod els > En er gy

The speed of sound under SSL condit ions is 340 m/ s so t hat our freest ream Mach

number is around 0. 15. This is low enough t hat we' ll assume t hat t he flow is

incompressible. So t he energy equat ion can be t urned off.

Make sure t here is no check in t he box next t o Ener gy Equat i on and click OK.

Defin e > Ma t er ia ls

Make sure ai r is select ed under Fl ui d Mat er i al s. Set Densi t y t o const ant and

equal t o 1.225 kg/ m

3

.

Click Change/ Cr eat e.

Defin e > Op er a t in g Con d it ion s

We' ll work in t erms of gauge pressures in t his example. So set Oper at i ng

Pr essur e t o t he ambient value of 101,325 Pa.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (4 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #4

Click OK.

Defin e > Bou n d a r y Con d it ion s

Set f ar f i el d1 and f ar f i el d2 t o t he vel oci t y - i nl et boundary t ype.

For each, click Set .... Then, choose Component s under Vel oci t y Speci f i cat i on

Met hod and set t he x- and y- component s t o t hat for t he freest ream. For

inst ance, t he x- component is 50* cos( 5

o

) = 49. 81.

Click OK.

Set f ar f i el d3 t o pr essur e- out l et boundary t ype, click Set ... and set t he Gauge

Pr essur e at t his boundary t o 0. Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (5 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #4

Go t o St ep 5: Solve!

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step4.htm (6 of 6)11/7/2005 6:56:31 PM

Flow over an Airfoil - Step #5

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Sol ve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Problem 2

Step 5: Solve!

Solve > Con t r ol > Solu t ion

Take a look at t he opt ions available.

Under Di scr et i zat i on, set Pr essur e t o PRESTO! and Moment um t o Second-

Or der Upw i nd.

Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (1 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #5

Solve > I n it ia lize > I n it ia lize. . .

As you may recall from t he previous t ut orials, t his is where we set t he init ial

guess values ( t he base case) for t he it erat ive solut ion. Once again, we' ll set

t hese values t o be t he ones at t he inlet . Select f ar f i el d1 under Comput e Fr om.

Click I ni t .

Solve > Mon it or s > Res id u a l. . .

Now we will set t he residual values ( t he crit eria for a good enough solut ion) .

Once again, we' ll set t his value t o 1e- 06.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (2 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #5

Click OK.

Solve > Mon it or s > For ce. . .

Under Coef f i ci ent , choose Li f t . Under Opt i ons, select Pr i nt and Pl ot . Then,

Choose ai r f oi l under Wal l Zones.

Last ly, set t he For ce Vect or component s for t he lift . The lift is t he force

perpendicular t o t he direct ion of t he freest ream. So t o get t he lift coefficient , set

X t o - sin( 5°) = -0.0872 and Y t o cos( 5°) = 0.9962.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (3 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #5

Click Appl y for t hese changes t o t ake effect .

Similarly, set t he For ce Moni t or opt ions for t he Dr ag force. The drag is defined

as t he force component in t he direct ion of t he freest ream. So under For ce

Vect or , set X t o cos( 5°) = 0.9962 and Y t o sin( 5°) = 0.0872. Turn on only Pr i nt

for it .

Rep or t > Refer en ce Va lu es

Now, set t he reference values t o set t he base cases for our it erat ion. Select

f ar f i el d1 under Comput e Fr om.

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (4 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #5

Click OK.

Ma in Men u > File > Wr it e > Ca s e. . .

Save t he case file before you st art t he it erat ions.

Solve > I t er a t e

What does t he convergence plot look like?

How many it erat ions does it t ake t o converge?

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (5 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #5

Ma in Men u > File > Wr it e > Ca s e & Da t a . . .

Save case and dat a aft er you have obt ained a converged solut ion.

Go t o St ep 6: Analyze Result s

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step5.htm (6 of 6)11/7/2005 6:56:50 PM

Flow over an Airfoil - Step #6

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Anal y ze Resul t s

7. Refine Mesh

Problem 1

Problem 2

Step 6: Analyze Results

Pl ot Pr essur e Coef f i ci ent

Plot > XY Plot . . .

Change t he Y Ax i s Funct i on t o Pr essur e..., followed by Pr essur e Coef f i ci ent .

Then, select ai r f oi l under Sur f aces.

Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step6.htm (1 of 3)11/7/2005 6:57:27 PM

Flow over an Airfoil - Step #6

( Click pict ure for larger image)

Pl ot Pr essur e Cont our s

Plot st at ic pressure cont ours.

Dis p la y > Con t ou r s . . .

Select Pr essur e... and St at i c Pr essur e from under Cont our s Of . Click Di spl ay .

( Click pict ure for larger image)

Where are t he highest and lowest pressures occurring?

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step6.htm (2 of 3)11/7/2005 6:57:27 PM

Flow over an Airfoil - Step #6

Go t o St ep 7: Refine Mesh

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step6.htm (3 of 3)11/7/2005 6:57:27 PM

Flow over an Airfoil - Step #7

Fl ow over an

Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Ref i ne Mesh

Problem 1

Problem 2

Step 7: Refine Mesh

* * Under const ruct ion* *

Go t o Problem 1

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/step7.htm11/7/2005 6:57:53 PM

Flow over an Airfoil - Problem #1

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Pr obl em 1

Problem 2

Problem 1

Consider t he incompressible, inviscid airfoil calculat ion in FLUENT present ed in

class. Recall t hat t he angle of at t ack, α, was 5°.

Repeat t he calculat ion for t he airfoil for α = 0° and α = 10°. Save your

calculat ion for each angle of at t ack as a different case file.

( a) Graph t he pressure coefficient ( C

p

) dist ribut ion along t he airfoil surface at α

= 5° and α = 10° in t he manner discussed in class ( i. e. , follow t he aeronaut ical

convent ion of let t ing C

p

decrease wit h increasing ordinat e ( y- axis) values) .

What change do you see in t he C

p

dist ribut ion on t he upper and lower surfaces

as you increase t he angle of at t ack?

Which part of t he airfoil surface cont ribut es most t o t he increase in lift wit h

increasing α?

Hi nt : The area under t he C

p

vs. x curve is approximat ely equal t o C

l

.

( b) Make a t able of C

l

and C

d

values obt ained for α = 0°, 5°, and 10°. Plot C

l

vs.

α for t he t hree values of α. Make a linear least - squares fit of t his dat a and obt ain

t he slope. Compare your result t o t hat obt ained from inviscid, t hin- airfoil t heory:

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/ps1.htm (1 of 2)11/7/2005 6:58:21 PM

Flow over an Airfoil - Problem #1

,

where α is in degrees.

Go t o Problem 2

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/ps1.htm (2 of 2)11/7/2005 6:58:21 PM

Flow over an Airfoil - Problem #2

Fl ow over an Ai r f oi l

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Problem 1

Pr obl em 2

Problem 2

Repeat t he incompressible calculat ion at α = 5° including viscous effect s. Since

t he Reynolds number is high, we expect t he flow t o be t urbulent . Use t he k- ε

t urbulence model wit h t he enhanced wall t reat ment opt ion. At t he farfield

boundaries, set t urbulence int ensit y= 1% and t urbulent lengt h scale= 0. 01.

( a) Graph t he pressure coefficient ( C

p

) dist ribut ion along t he airfoil surface for

t his calculat ion and t he inviscid calculat ion done in t he previous problem at α =

5°. Comment on any differences you observe.

( b) Compare t he C

l

and C

d

values obt ained wit h t he corresponding values from

t he inviscid calculat ion. Discuss briefly t he similarit ies and differences bet ween

t he t wo result s.

Back t o Problem Specificat ion

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/ps2.htm11/7/2005 6:58:29 PM

Fluent - Forced Convection on a Flat Plate Step #1

For ced Convect i on over a Fl at Pl at e

Problem Specificat ion

1. Cr eat e Geomet r y i n GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Step 1: Create Geometry in GAMBIT

St ar t GAMBI T & Sel ect Sol ver

Specify t hat t he mesh t o be creat ed is for use wit h FLUENT 6:

Ma in Men u > Solver > FLUENT 5/ 6

Verify t his has been done by looking in t he Transcript Window where you should

see:

The boundary t ypes t hat you' ll be able t o select in t he t hird st ep depends on

t he solver select ed.

St r at egy f or cr eat i ng f l ow f i el d geomet r y

I n creat ing t he geomet ry for our flow field we must consider what is necessary

for our model t o approximat e real flow. A boundary layer grows along t he plat e,

which must sat isfy t he no slip condit ion. The flow velocit y at t he plat e must be

zero. Cont inuit y requires t hat t his condit ion gives rise t o a y- velocit y. Alt hough

t he y- velocit y is significant ly smaller in magnit ude t han t he x- velocit y, it can

affect t he solut ion significant ly if not t aken int o considerat ion when creat ing t he

geomet ry of t he flow field.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (1 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

We will put t he origin of t he coordinat e syst em at t he lower left corner of t he

rect angle t hat defines our flow field. The coordinat es of t he corners are shown in

t he figure below:

We will first creat e four vert ices at t he four corners and j oin adj acent vert ices t o

get t he edges of t he rect angle. We will t hen form a face t hat covers t he area of

t he rect angle.

Cr eat e Ver t i ces

We will t reat t his problem as a 2- dimensional problem by assuming t hat t he

plat e is infinit ely wide. Let ' s begin by creat ing t he vert ices t hat define our flow

region.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ver t ex Com m a n d

Bu t t on > Cr ea t e Ver t ex

Not e t hat t he Creat e Vert ex but t on has already been select ed by default . Aft er

you select a but t on under a sub- pad, it becomes t he default when you go t o a

different sub- pad and t hen come back t o t he sub- pad.

Creat e t he vert ex at t he lower- left corner of t he rect angle:

Next t o x : , ent er value 0. Next t o y : , ent er value 0. Next t o z: , ent er value 0

( t hese values should be default s) . Click Appl y .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (2 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

This creat es t he vert ex ( 0, 0, 0) which is displayed in t he graphics window.

I n t he Transcript window, GAMBI T report s t hat it "Creat ed vert ex: vert ex. 1".

The vert ices are numbered vert ex. 1, vert ex. 2 et c. in t he order in which t hey

are creat ed.

Repeat t his process t o creat e t hree more vert ices:

Vert ex 2: ( 1, 0, 0)

Vert ex 3: ( 1, 1, 0)

Vert ex 4: ( 0, 1, 0)

Not e t hat for a 2D problem, t he z- coordinat e can always be left t o t he default

value of 0.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (3 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

Op er a t ion Toolp a d > Globa l Con t r ol > Fit t o Win d ow Bu t t on

This fit s t he four vert ices of t he rect angle we have creat ed t o t he size of t he

Graphics Window.

( click pict ure for larger image)

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (4 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

Anot her useful but t on on t he Op er a t ion Toolp a d is t he Or ien t Mod el but t on .

I f you click and hold t he left mouse but t on and t hen move t he mouse, t he model

will rot at e 3- dimensionally. This is, of course, not usually a helpful feat ure when

creat ing 2- D models in GAMBI T. Click t he Or ien t Mod el but t on t o make t he z-

axis normal t o t he page again.

Cr eat e Edges

An edge is creat ed by select ing t wo vert ices and creat ing a line bet ween t hem.

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on

> Cr ea t e Ed ge

Click t he up ar r ow but t on next t o t he vert ices box in t he Creat e St raight

Edge window.

This brings up a list of vert ices, from which vert ices 1 and 2 can be select ed.

Select Ver t ex .1 and Ver t ex .2. Then push t he r i ght ar r ow but t on t o

bring t hese vert ices int o t he Picked column.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (5 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

Click Cl ose. Then click Appl y in t he Creat e St raight Edge window t o creat e t his

edge.

Alt ernat ely, t hese vert ices can be select ed by holding down t he Shi f t but t on and

clicking on t he corresponding vert ices. As each vert ex is picked, it will appear red

in t he Graphics Window. Then let go of t he Shi f t but t on and click Appl y in t he

Creat e St raight Edge window.

Repeat t his process t o creat e edges bet ween vert ices 2 & 3, vert ices 3 & 4, and

vert ices 4 & 1.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (6 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

( click pict ure for larger image)

Cr eat e Face

Op er a t ion Toolp a d > Geom et r y Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on

> For m Fa ce

To form a face out of t he area enclosed by t he four lines, we need t o select t he

four edges t hat enclose t his area. This is done in much t he same way as when

we select ed t he vert ices.

Click t he up ar r ow but t on next t o t he vert ices box in t he Creat e Face From

Wireframe window. Then push t he Al l r i ght ar r ow but t on t o bring t hese

vert ices int o t he Picked column.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (7 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

Click Cl ose. Then click Appl y in t he Creat e Face From Wireframe window t o

creat e t he face. The edges and vert ices will become blue, indicat ing t hat t hey

now form a face.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (8 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

( click pict ure for larger image)

Sav e

Save your GAMBI T file in your working direct ory.

Ma in Men u > File > Sa ve As . . . > Br ows e. . .

Find your working direct ory and save your GAMBI T file t here. Make sure t o ent er

t he file name, plate.dbs, in t he Select ion box in addit ion t o t he pat h.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (9 of 10)11/7/2005 6:59:38 PM

Fluent - Forced Convection on a Flat Plate Step #1

Go t o St ep 2: Mesh Geomet ry in GAMBI T

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step1.htm (10 of 10)11/7/2005 6:59:38 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #2

For ced Convect i on over a Fl at Pl at e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet r y i n GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Step 2: Mesh Geometry in GAMBIT

We' ll now creat e a mesh on t he rect angular face wit h 100 divisions in t he vert ical

direct ion and 30 divisions in t he horizont al direct ion. We' ll first mesh t he four

edges and t hen t he face. The desired grid spacing is specified t hrough t he edge

mesh.

Mesh Edges

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Ed ge Com m a n d Bu t t on >

Mes h Ed ges

Mesh St r at egy

I n creat ing t his mesh, it is desirable t o have more cells near t he plat e ( Edge 1)

because we want t o resolve t he t urbulent boundary layer, which is very t hin

compared t o t he height of t he flow field.

Click t he up ar r ow but t on next t o t he Edges box in t he Mesh Edges window.

Select edge Edge.2. Then push t he r i ght ar r ow but t on t o bring t his

vert ex int o t he Picked column. Not ice t hat t he arrow on t he select ed edge should

be point ing upwards. An upwards point ing arrow indicat es t he direct ion of closely

spaced nodes t o widely spaced nodes. Remember, we will need more closely

spaced nodes near t he boundary layer in order t o resolve it accurat ely.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step2.htm (1 of 4)11/7/2005 7:00:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #2

The proper arrow direct ion is necessary t o ensure a proper mesh. Select Edge.4

in t he Mesh Edges window. The arrow on t his edge is point ing downwards, which

needs t o be changed. Shi f t + Mi ddl e- cl i ck on t he select ed edge t o change t he

direct ion of t he arrow t o upward.

Under Type, select Successi ve Rat i o, if it is not already select ed. Set Rat io t o

1.08. Under Spacing, select I nt er val Count . Set I nt er val Count t o 100 and

t hen click Appl y .

Select Edge.1 and Edge.3 in t he Mesh Edges Window. The direct ion of t he

arrows on t hese edges is irrelevant because t he divisions will be t he same

lengt h. Leave t he Successi ve Rat i o set t o 1 and set t he I nt er val Count t o 30.

Click Appl y .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step2.htm (2 of 4)11/7/2005 7:00:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #2

( click pict ure for larger image)

Mesh Face

Op er a t ion Toolp a d > Mes h Com m a n d Bu t t on > Fa ce Com m a n d Bu t t on >

Mes h Fa ces

Shi f t l ef t - cl i ck on t he face or use t he up arrow next t o Faces t o select t he face.

Click Appl y .

( click pict ure for larger image)

Go t o St ep 3: Specify Boundary Types in GAMBI T

http://instruct1.cit.cornell.edu/courses/fluent/plate/step2.htm (3 of 4)11/7/2005 7:00:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #2

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step2.htm (4 of 4)11/7/2005 7:00:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #3

For ced Convect i on over a Fl at Pl at e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Speci f y Boundar y Ty pes i n GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Step 3: Specify Boundary Types in GAMBIT

Cr eat e Boundar y Ty pes

We' ll next set t he boundary t ypes in GAMBI T. The left edge is t he inflow of t he

flow field, t he right edge t he out flow, t he t op edge t he open t op of t he flow field,

and t he bot t om edge t he plat e.

Op er a t ion Toolp a d > Zon es Com m a n d Bu t t on > Sp ecify Bou n d a r y Typ es

http://instruct1.cit.cornell.edu/courses/fluent/plate/step3.htm (1 of 4)11/7/2005 7:01:52 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #3

Com m a n d Bu t t on

This will bring up t he Specify Boundary Types window on t he Operat ion Panel.

We will first specify t hat t he left edge is t he inflow. Under Ent it y: , pick Edges so

t hat GAMBI T knows we want t o pick an edge ( face is default ) .

Now select t he left edge by Shi f t - cl i ck i ng on it . The select ed edge should

appear in t he yellow box next t o t he Edges box as well as t he Label / Ty pe list

under t he Edges box.

Next t o Name: , ent er inflow.

For Type: , select VELOCI TY_I NLET. You may have t o move t he Specify

Boundary Types box up in order t o see t he bot t om of t he list and select

VELOCI TY_I NLET.

Click Appl y . You should see t he new ent ry appear under Name/ Ty pe box near

t he t op of t he window.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step3.htm (2 of 4)11/7/2005 7:01:52 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #3

Repeat t his process for t he ot her t hree edges according t o t he following t able:

Edge

Position

Name Type

Left inflow VELOCITY_INLET

Right outflow PRESSURE_OUTLET

http://instruct1.cit.cornell.edu/courses/fluent/plate/step3.htm (3 of 4)11/7/2005 7:01:52 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #3

Top top SYMMETRY

Bottom plate WALL

You should have t he following edges in t he Name/ Type list when finished:

Sav e and Ex por t

Ma in Men u > File > Sa ve

Ma in Men u > File > Exp or t > Mes h . . .

Type in plate.msh for t he Fi l e Name: . Select Ex por t 2d Mesh because t his is a

2 dimensional mesh. Click Accept .

I t is import ant t o check t hat plate.msh has been creat ed in your working

direct ory. GAMBI T may periodically fail t o writ e t he .msh file. I f t his should

happen, simply t ry writ ing t he .msh file t o anot her direct ory and t hen coping it

int o your working direct ory.

Go t o St ep 4: Set Up Problem in FLUENT

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step3.htm (4 of 4)11/7/2005 7:01:52 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

For ced Conv ect i on over a Fl at Pl at e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Pr obl em i n FLUENT

5. Solve!

6. Analyze Result s

7. Refine Mesh

Step 4: Set Up Problem in FLUENT

Launch Fl uent 6.0

St a r t > Pr ogr a m s > Flu en t I n c > FLUENT 6. 0

Select t he 2ddp version and click Run.

The "2ddp" opt ion is used t o select t he 2- dimensional, double- precision solver. I n t he double-

precision solver, each float ing point number is represent ed using 64 bit s in cont rast t o t he

single- precision solver which uses 32 bit s. The ext ra bit s increase not only t he precision but

also t he range of magnit udes t hat can be represent ed. The downside of using double

precision is t hat it requires more memory.

I mpor t Gr i d

Ma in Men u > File > Rea d > Ca s e. . .

Navigat e t o t he working direct ory and select t he plate.msh file. This is t he mesh file t hat was

creat ed using t he preprocessor GAMBI T in t he previous st ep. FLUENT report s t he mesh

st at ist ics as it reads in t he mesh:

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (1 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Check t he number of nodes, faces ( of different t ypes) and cells. There are 3000 quadrilat eral

cells in t his case. This is what we expect because we used 30 divisions in t he horizont al

direct ion and 100 divisions in t he vert ical direct ion while generat ing t he grid. So t he t ot al

number of cells is 30* 100 = 3000.

Also, t ake a look under zones. We can see t he four zones i nf l ow , out f l ow , t op, and pl at e t hat

we defined in GAMBI T.

Check and Di spl ay Gr i d

First , we check t he grid t o make sure t hat t here are no errors.

Ma in Men u > Gr id > Ch eck

Any errors in t he grid would be report ed at t his t ime. Check t he out put and make sure t hat

t here are no errors report ed. Check t he grid size:

Ma in Men u > Gr id > I n fo > Size

The following st at ist ics should appear:

Display t he grid:

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (2 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Ma in Men u > Dis p la y > Gr id . . .

Make sure all 5 it ems under Surfaces are select ed.

Then click Di spl ay . The graphics window opens and t he grid is displayed in it . Your grid should

look like t his:

( click pict ure for larger image)

Def i ne Sol v er Pr oper t i es

Ma in Men u > Defin e > Mod els > Solver

We' ll use t he default s of 2D space, segregat ed solver, implicit formulat ion, st eady flow and

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (3 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

absolut e velocit y formulat ion. Click OK.

Ma in Men u > Defin e > Mod els > En er gy

We are int erest ed in solving t he t emperat ure dist ribut ion, so we need t o solve t he energy

equat ion. Select t he Ener gy Equat i on and click OK t o exit t he menu.

Ma in Men u > Defin e > Mod els > Vis cou s

Under Model, select t he k - epsi l on t urbulence model. We will use t he Real i zabl e model in t he

k- epsilon Model box. The Realizable k- epsilon model produces more accurat e result s for

boundary layer flows t han t he St andard k- epsilon model. I n t he Near- Wall Treat ment box,

observe t he Enhanced Wal l Tr eat ment opt ion, which deals wit h t he resolut ion of t he boundar

layer in our model. There are 3 regions in t he boundary layer t hat we are concerned wit h,

st art ing at t he wall:

1. Laminar sublayer ( y+ < 5)

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (4 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

2. Buffer region ( 5 < y+ < 30)

3. Turbulent region ( y+ > 30)

y+ is a mesh- dependent dimensionless dist ance t hat quant ifies t o what degree t he wall layer is

resolved. Aft er solving t his problem in FLUENT, we will observe t he value of y+ for each mesh

we use. The Enhanced Wal l Tr eat ment opt ion serves t o more accurat ely resolve t he

boundary layer in t he case when t he mesh is only fine enough t o resolve t o t he t urbulent region

( y+ > 30) . Enhanced Wal l Tr eat ment also improves t he accuracy of meshes t hat can only be

resolved t o t he Buffer region ( 5< y+ < 30) . However, solut ions wit h y+ values in t he buffer

region are generally less accurat e t han if t he solut ion is resolved t o one of t he ot her 2 regions.

Look at FLUENT Help sect ion 10. 9, Grid Considerat ions for Turbulent Flow Simulat ions, for more

det ails.

For our mesh, FLUENT will be able t o resolve t he laminar sublayer, t hus Enhanced Wal l

Tr eat ment does not improve t he accuracy of our solut ion wit h our mesh. I t will however make

a difference in St ep 7 when we use a less refined mesh. The t hickness of t he boundary layer is

significant ly smaller t han t he height of our flow field. Resolving t he solut ion t o t he laminar

sublayer is comput at ionally int ensive, especially in high Reynolds Number flows. Resolving t o

t he t urbulent region is oft en t he only reasonable opt ion. Thus it is good pract ice t o always use

Enhanced Wal l Tr eat ment when dealing wit h a boundary layer. Alt hough it is not necessary

wit h t he current mesh, it will be necessary for t he less refined mesh lat er on, so go ahead and

select Enhanced Wal l Tr eat ment now.

Select Ther mal Ef f ect s in t he Enhanced Wall Treat ment Opt ions box t o include t he t hermal

t erms in t he Enhanced Wall Treat ment equat ion.

The values in t he Model Const ant s box are const ant s used in t he k - epsi l on t urbulence

equat ions. These values for t he Model Const ant s are well- accept ed for a wide range of wall-

bounded shear flows. Leave all values in t he Model Const ant s box set t o t heir default values.

Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (5 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Def i ne Mat er i al Pr oper t i es

Ma in Men u > Defin e > Ma t er ia ls . . .

Change Densi t y t o ideal gas because we are t reat ing t he flow as compressible. FLUENT will

calcualt e t he densit y of t he flow at each point based on t he pressure and t emperat ure it

calculat es at t hat point . Leave Cp set as t he default value of 1006.43. Change Ther mal

Conduct i vi t y t o 9.4505 e-4. Change Vi scosi t y t o 6.667e-7. Scroll down t o see Mol ecul ar

Wei ght . Leave Mol ecul ar Wei ght set t o t he default value of 28.966. These are t he values

t hat we specified under Problem Specificat ion.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (6 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Click Change/ Cr eat e. Simply clicking close wit hout clicking Change/ Creat e will cause t hese

propert ies t o revert back t o t heir default values.

Def i ne Oper at i ng Condi t i ons

Ma in Men u > Defin e > Op er a t in g Con d it ion s . . .

For all flows, FLUENT uses gauge pressure int ernally. Any t ime an absolut e pressure is needed,

it is generat ed by adding t he operat ing pressure t o t he gauge pressure. We' ll use t he default

value of 1 at m ( 101, 325 Pa) as t he Oper at i ng Pr essur e.

Click Cancel t o leave t he default value in place.

Def i ne Boundar y Condi t i ons

We' ll now set t he value of t he velocit y at t he inflow and pressure at t he out flow.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (7 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Ma in Men u > Defin e > Bou n d a r y Con d it ion s . . .

We not e here t hat t he four t ypes of boundaries we defined are specified as zones on t he left

side of t he Boundary Condit ions Window. There are also 2 zones default - int erior fluid, used t o

define t he int erior of t he flow field. We will not need t o change any set t ing for t hese 2 zones.

Move down t he list and select i nf l ow under Zone. Not e t hat FLUENT indicat es t hat t he Type of

t his boundary is vel oci t y - i nl et . Recall t hat t he boundary t ype for t he inflow was set in GAMBI T.

I f necessary, we can change t he boundary t ype set previously in GAMBI T in t his menu by

select ing a different t ype from t he list on t he right . Click Set ....

Ent er 1 for Vel oci t y Magni t ude. This set s t he velocit y of t he fluid ent ering at t he left boundary

t o a uniform velocit y profile of 1m/ s. Set Temperat ure t o 353K. Change Turbulence Specificat ion

Met hod t o I nt ensi t y and Vi scosi t y Rat i o. Set Turbulence I nt ensit y t o 1 and Turbulent

Viscosit y Rat io t o 1. Click OK.

Choose out f l ow under Zone. The Type of t his boundary is pr essur e- out l et . Click Set .... The

default value of t he Gauge Pr essur e is 0. The ( absolut e) pressure at t he out flow is 1 at m.

Since t he operat ing pressure is set t o 1 at m, t he out flow gauge pressure = out flow absolut e

pressure - operat ing pressure = 0. Because we do not expect any backflow, we do not need t o

set any backflow condit ions. Click Cancel t o leave t he default s in place.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (8 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

Click on pl at e under Zones and make sure Type is set as w al l . Click Set .... Because we have a

heat ed isot hermal plat e, we need t o set t he t emperat ure. On t he Ther mal t ab, select

Temper at ur e under Thermal Condit ions. Change Temperat ure t o 413. The mat erial select ed is

inconsequent ial because t he plat e has zero t hickness in our model, t hus t he mat erial propert ies

of t he plat e do not affect t he heat t ransfer propert ies of t he plat e. Click OK.

The last boundary condit ion t o set is for t he t op of t he flow field. Click on t op under Zones and

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (9 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #4

make sure Type is set as sy mmet r y . Click Set ... t o see t hat t here is not hing t o set for t his

boundary. Click OK.

Click Cl ose t o close t he Boundary Condit ions menu.

Go t o St ep 5: Solve!

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step4.htm (10 of 10)11/7/2005 7:04:00 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

For ced Conv ect i on ov er a Fl at Pl at e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Sol ve!

6. Analyze Result s

7. Refine Mesh

Step 5: Solve!

We' ll use a second- order discret izat ion scheme.

Ma in Men u > Solve > Con t r ols > Solu t ion . . .

Change Densi t y , Moment um, Tur bul ence Ki net i c Ener gy , Tur bul ence Di ssi pat i on

Rat e, and Ener gy all t o Second Or der Upw i nd. Leave Pressure and Pressure- Velocit y

Coupling set t o t he default met hods ( St andar d and SI MPLE, respect ively) . The ot her

Pressure and Pressure- Velocit y Coupling met hods are useful for flows wit h part icular

charact erist ics not present in our problem.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (1 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

Click OK.

Set I ni t i al Guess

I nit ialize t he flow field t o t he values at t he inflow:

Ma in Men u > Solve > I n it ia lize > I n it ia lize. . .

I n t he Solut ion I nit ializat ion window t hat comes up, choose i nf l ow under Comput e

From. The X Vel oci t y for all cells will aut omat ically be set t o 1 m/ s, t he Y Vel oci t y t o 0

m/ s and t he Gauge Pr essur e t o 0 Pa. These values have been t aken from t he inflow

boundary condit ion.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (2 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

Click I ni t . This complet es t he init ializat ion. Then click Cl ose.

Set Conver gence Cr i t er i a

FLUENT report s a residual for each governing equat ion being solved. The residual is a

measure of how well t he current solut ion sat isfies t he discret e form of each governing

equat ion. We will it erat e unt il t he residual for each equat ion falls below 1e- 6.

Ma in Men u > Solve > Mon it or s > Res id u a l. . .

Change t he residual under Conver gence Cr i t er i on for cont i nui t y , x - vel oci t y , and y -

vel oci t y , ener gy , k , and epsi l on all t o 1e- 6.

Also, under Opt i ons, select Pr i nt and Pl ot . This will print t he residuals in t he main

window and plot t he residuals in t he graphics window as t hey are calculat ed.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (3 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

Click OK.

This complet es t he problem specificat ion. Save your work:

Ma in Men u > File > Wr it e > Ca s e. . .

Type in plate.cas for Case Fi l e. Click OK. Check t hat t he file has been creat ed in your

working direct ory. I f you exit FLUENT now, you can ret rieve all your work at any t ime

by reading in t his case file.

I t er at e Unt i l Conver gence

St art t he calculat ion by running 10, 000 it erat ions. The solut ion will converge before

10, 000 it erat ions are performed, which will st op t he it erat ion process.

Ma in Men u > Solve > I t er a t e. . .

I n t he I t erat e Window, change t he Number of I t er at i ons t o 10000. Click I t er at e.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (4 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

The residuals for each it erat ion are print ed out as well as plot t ed in t he graphics window

as t hey are calculat ed.

( Click pict ure for larger image)

The residuals fall below t he specified convergence crit erion of 1e- 6 in approximat ely

1623 it erat ions.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (5 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #5

Save t he solut ion t o a dat a file:

Ma in Men u > File > Wr it e > Da t a . . .

Ent er plate.dat for Dat a Fi l e and click OK. Check t hat t he file has been creat ed in

your working direct ory. You can ret rieve t he current solut ion from t his dat a file at any

t ime.

Go t o St ep 6: Analyze Result s

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step5.htm (6 of 6)11/7/2005 7:05:42 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

For ced Convect i on over a Fl at Pl at e

Problem Specificat ion

1. Creat e Geomet ry in GAMBI T

2. Mesh Geomet ry in GAMBI T

3. Specify Boundary Types in GAMBI T

4. Set Up Problem in FLUENT

5. Solve!

6. Anal y ze Resul t s

7. Refine Mesh

Step 6: Analyze Results

y +

Turbulent flows are significant ly affect ed by t he presence of walls. The k- epsilon

t urbulence model' s validit y is grid- independent away from walls but requires

verificat ion t o make sure it is valid when used near walls. The near- wall model is

sensit ive t o t he grid resolut ion, which is assessed in t he wall unit y+ , as

discussed in St ep 4.

First , we need t o set t he reference values needed t o calculat e y+ .

Ma in Men u > Rep or t > Refer en ce Va lu es . . .

Select i nf l ow under Comput e Fr om t o t ell FLUENT t o use values at t he inflow

for t he reference values. Check t hat t he reference value for velocit y is 1 m/ s,

t emperat ure is 353 K, and coefficient of viscosit y is 6. 667e- 7 kg/ m- s as given in

t he Problem Specificat ion. These reference values will be used t o non-

dimensionalize t he dist ance of t he cell cent er from t he wall t o obt ain t he

corresponding y+ values. Click OK.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (1 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

By using t he following met hod, plot y+ values for wall- adj acent cells t o check

how t hey compare wit h t he recommendat ion ment ioned above.

Ma in Men u > Plot > XY Plot . . .

Make sure t hat Posi t i on on X Ax i s is set under Opt i ons, t hat 1 is t he value

next t o X, and 0 is t he value next t o Y under Pl ot Di r ect i on. Recall t hat t his

t ells FLUENT t o plot t he x- coordinat e value on t he abscissa of t he graph. Select

Tur bul ence... under Y Ax i s Funct i on and select Wal l Ypl us from t he drop

down list under t hat . Since we want t he y+ value for cells adj acent t o t he wall of

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (2 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

t he pipe, choose pl at e under Sur f aces.

Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (3 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

( Click pict ure for larger image)

As we can see, t he wall y+ value is bet ween 1. 0 and 1. 4 ( ignoring t he

anamolous at t he inflow) . Because t hese values are less t han 5, t he near- wall

mesh resolut ion is in t he laminar sublayer, which is t he most accurat e region t o

which we can resolve t he boundary layer.

Sav e Pl ot

I n t he Solut ion XY Plot Window, check t he Wr i t e t o Fi l e box under Opt i ons. The

Pl ot but t on should have changed t o t he Wr i t e... but t on. Click on Wr i t e.... Ent er

yplus.xy as t he filename and click OK. Check t hat t his file has been creat ed in

your FLUENT working direct ory.

Vel oci t y at x = 1m

Ma in Men u > Plot > XY Plot . . .

Under Opt ions, unselect Posi t i on on X Ax i s and select Posi t i on on Y Ax i s.

Under Plot Direct ion, ent er 0 in t he X box and 1 in t he Y box. This t ells FLUENT

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (4 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

t o plot a vert ical rat her t han horizont al profile.

Under X Ax i s Funct i on, pick Vel oci t y ... and t hen in t he box under t hat , pick X

Vel oci t y . Finally, select out f l ow under Sur f aces since we are plot t ing t he

velocit y profile at t he out flow. De- select pl at e under Sur f aces.

Click on Ax es... in t he Solut ion XY Plot window. Select X in t he Axis box. I n t he

Opt ions box select Maj or Rul es t o t urn on t he grid lines in t he plot . Click Appl y .

Then select t he Y in t he Axis box, select Maj or Rul es again, and t urn off Aut o

Range. I n t he Range box ent er 0.1 for t he Maximum so t hat we may view t he

velocit y profile in t he boundary layer region more closely. Click Appl y and

Cl ose.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (5 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

Uncheck Wr i t e t o Fi l e. Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (6 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

( Click pict ure for larger image)

We not ice here t hat t he x velocit y reaches 1 m/ s at approximat ely y = 0. 02 m.

This shows t he relat ive t hinness of t he boundary layer compared t o t he lengt h

scale of t he plat e. We also not ice t hat t he velocit y profile is slight ly great er t han

1 m/ s above t he boundary layer. We know t his would not happen in real flow,

rat her it is a result of t he boundary condit ion we have chosen for our model. We

chose t he Sy mmet r y boundary condit ion at t he t op of our flow field, which is

essent ially a wall wit hout t he no- slip condit ion. Thus, no flow is permit t ed t o

escape t hrough t his boundary.

I n a real ext ernal flow, t here is no such boundary at t he t op and flow is

permit t ed t o pass t hrough freely. When we consider t he inflow and out flow

velocit y profiles in t erms of conservat ion of mass, t he uniform velocit y profile of

1 m/ s at x = 0 has more mass ent ering t he flow field t han t he non- uniform

velocit y profile at x = 1m, in which t he velocit y is lower near t he plat e. I n

addit ion, t he fluid is expanding near t he plat e because it s t emperat ure is

increasing, furt her increasing t he y- velocit y of t he fluid above it . These fact ors

require t hat some mass must escape t hrough t he t op of our flow field in order t o

sat isfy conservat ion of mass.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (7 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

Choosing a Pr essur e Out l et for t he t op boundary condit ion would represent real

ext ernal flow more accurat ely. Unfort unat ely, t his cannot be used in our flow

field wit hout encount ering convergence problems, so select ing t he Sy mmet r y

boundary condit ion was t he next best opt ion. Because we are not allowing flow

t o escape t hrough t he t op boundary, we observe an out flow velocit y profile in

which out flow velocit y is great er t han 1 above t he boundary layer in order t o

sat isfy conservat ion of mass. Fort unat ely, t he inaccuracies result ing from t he

model we chose have no significant effect on t he heat t ransfer coefficient s at t he

plat e.

Select Wr i t e t o Fi l e and save t he dat a for t his plot as outflow_profile.xy.

Pl ot Nussel t Number vs. Rey nol ds Number

Recall t hat t he Nusselt Number is a non- dimensional heat t ransfer coefficient t hat

relat es convect ive and conduct ive heat t ransfer.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (8 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

I n order t o obt ain t he Nusselt Number from FLUENT, we will begin by plot t ing

Tot al Surface Heat Flux.

Ma in Men u > Plot > XY Plot . . .

I n t he Opt ions box, change back t o Posi t i on on X Ax i s. I n t he Plot Direct ion

box, ent er t he default values of 1 in t he X box and 0 in t he Y box. Under Y- Axis

Funct ion choose Wal l Fl ux es. I n t he box below, chose Tot al Sur f ace Heat

Fl ux . Select Pl at e under Surfaces. Before plot t ing, be sure t o t urn on Aut o

Range for t he Y axis under Ax es....

Click Pl ot .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (9 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

( click pict ure for larger image)

Now Select Wr i t e t o Fi l e. Save t he dat a for t his plot as heatflux.xy. Click

Wr i t e....

Open t he file heatflux.xy using Wordpad or a similar applicat ion. You can

simply copy and past e t he dat a int o Excel.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (10 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

I f Excel does not aut omat ically separat e t he dat a int o columns, separat e it by

select ing t he column of dat a and t hen using t he Text t o Columns funct ion:

Ma in Men u > Da t a > Text t o Colu m n s

The first column is t he x locat ion on t he plat e and t he second column is t he t ot al

surface heat flux ( q' ' ) at t he corresponding x locat ion. We now need t o det ermine

t he Nusselt number from t hese values at each x locat ion. We will define posit ive

q' ' as heat t ransfer int o t he fluid. Use t he following expression t o convert q' ' t o

Nusselt Number in your Excel spreadsheet .

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (11 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

Reynolds Number can be defined at each x locat ion by

Now plot Re vs. Nu in Excel. Your plot should look like t his:

( click pict ure for larger image)

Compar e Resul t s w i t h Cor r el at i on & Ex per i ment

Validat e your result s form FLUENT by comparing t o a correlat ion and

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (12 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

experiment al result s. The correlat ion we will use is derived by Reynolds [ 1] :

All propert ies in t his correlat ion are evaluat ed at t he free- st ream st at ic

t emperat ure of 300K. This correlat ion assumes t he following:

1. Pr = 0. 7

2. 10^ 5 < Re < 10^ 7

3. Fluid propert ies evaluat ed at free- st ream condit ions

4. Turbulent compressible boundary layer

5. Flat plat e

6. Frict ion fact or calculat ed from t he following relat ion ( implicit in Nu equat ion

above, does not need t o be calculat ed in your analysis) :

Add t he Reynolds correlat ion for Nusselt Number t o your Excel spreadsheet .

Seban & Dought y [ 2] performed a heat ed flat plat e experiment for which t hey

derived t he following expression for Nusselt Number:

The Seban & Dought yexperiment was performed wit h air as t he fluid ( Pr = 0. 7)

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (13 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

and at various Reynolds Numbers in t he range 1e5 < Re < 4e6. Add t he t his

experiment al relat ion for Nusselt Number t o your Excel spreadsheet .

Now plot and compare Re vs. Nu from FLUENT, t he Reynolds Correlat ion, and

Seban' s experiment .

( click pict ure for larger image)

As we can see, t here is very lit t le variat ion bet ween t hese 3 result s. The largest

% error bet ween t he FLUENT result s and t he Reynolds correlat ion is only 7. 5%.

I n t urbulent flow as we have here, similar result s bet ween FLUENT and

correlat ion are more difficult t o come by t han in laminar flow because a t urbulent

model must be used in FLUENT, which does not solve t he Navier- St okes

Equat ions exact ly. Experiment al error ( in experiment s from which correlat ions

are derived) also account s for some of t his 7. 5% error. Each of t he t urbulence

models t hat FLUENT offers produces result s similar t o t hese, alt hough t he k-

epsilon model is t he most appropriat e model t o use in t his case.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (14 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Forced Convection on a Flat Plate Step #6

Go t o St ep 7: Refine Mesh

[ 1] Reynolds, W. C. , Kays, W. M. , Kline, S. J. "Heat Transfer in t he Turbulent

I ncompressible Boundary Layer. " NASA Memo 12- 1- 58W. December 1958.

[ 2] Seban, R. A. and Dought y, D. L. "Heat Transfer t o Turbulent Boundary Layers

wit h Variable Freest ream Velocit y. " Journal of Heat Transfer 78: 217 ( 1956) .

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step6.htm (15 of 15)11/7/2005 7:07:43 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

For ced Conv ect i on ov er a Fl at Pl at e

Problem Specificat ion

1. St art - up and preliminary set - up

2. Creat e Geomet ry

3. Mesh Geomet ry

4. Set Up Problem in FLUENT

5. Solve!

6. Analyze Result s

7. Ref i ne Mesh

Step 7: Refine Mesh

I t is very import ant t o assess t he dependence of your result s on t he mesh used by

repeat ing t he same calculat ion on different meshes and comparing t he result s. We will

re- do t he previous calculat ion on a 30 x 50 mesh as well as a 30 x 150 mesh and t hen

compare t he result s wit h t he 30x100 mesh used previously.

Modi f y Mesh i n GAMBI T t o a 30x 50 mesh

The 30x100 mesh is saved as plat e. dbs in your working direct ory. Bring up t he

command prompt window as in st ep 1. To copy plat e. dbs t o plat e50. dbs, at t he

command prompt , t ype

copy plate.dbs plate50.dbs

We will work wit h plat e50. dbs in order t o ret ain plat e. dbs as is. Launch GAMBI T wit h

plat e50. dbs as t he input file by t yping:

gambit plate50.dbs

Follow t he same met hod as in previous t ut orials t o change t he mesh. The face mesh will

be aut omat ically delet ed when you re- mesh t he edges. The t op and bot t om edges will

remain t he same.

Mesh t he inflow and out flow edges at a Successi ve Rat i o of 1. 095 and an I nt er val

Count of 50.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (1 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Remesh t he face and t hen export t his as t he 2D mesh file, plate50.msh.

Read t he file int o FLUENT and repeat st ep 4 and st ep 5 of t his t ut orial t o set up and

solve t he problem in FLUENT. The solut ion should converge in approximat ely 115

it erat ions. Plot y+ at t he plat e as explained in st ep 6.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (2 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( click pict ure for larger image)

y+ ranges from 29 t o 50 in t his plot . This is ( most ly) out side of t he ill- defined Buffer

region ( 5 < y+ < 30) and is t hus accept able.

Now use t he Tot al Surface Heat Flux plot t o det ermine Nu( x) . Plot Re vs. Nu and

compare wit h t he 30x100 mesh result s.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (3 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( click pict ure for larger image)

We can see t hat t he courser mesh produces slight ly different result s, alt hough t hey are

st ill reasonable. Some numerical error is int roduced when t he less- refined 30x50 mesh

is used. As one would expect , resolving t he boundary layer t o t he laminar sublayer,

which we did wit h t he orignial mesh, produces more accurat e result s t han resolving only

t o t he t urbulent region. Resolving t o t he laminar sublayer is not always a reasonable

t hing t o do, especially at high Reynolds numbers. The result s from using t he 30 x 50

grid show t hat a reasonable solut ion can st ill be obt ained wit hout resolving down t o t he

laminar sublayer.

Modi f y Mesh i n GAMBI T t o a 30x 150 mesh

Creat e a mesh t hat is finer t han t he original mesh t o see if our original solut ion

cont ained inaccuracies due t o t he mesh. Mesh t he inflow and out flow edges at a

Successi ve Rat i o of 1. 065 and an I nt er val Count of 150.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (4 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

Remesh t he face and t hen export t his as t he 2D mesh file, plate150.msh.

Read t he file int o FLUENT and repeat st ep 4 and st ep 5 of t his t ut orial t o set up and

solve t he problem in FLUENT. The solut ion should converge in approximat ely 4550

it erat ions. Plot y+ at t he plat e.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (5 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( click pict ure for larger image)

y+ ranges from 0. 14 t o 0. 25 in t his plot , well wit hin t he laminar sublayer.

Now use t he Tot al Surface Heat Flux plot t o det ermine Nu( x) . Plot Re vs. Nu and

compare wit h t he 30x100 mesh result s.

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (6 of 7)11/7/2005 7:09:12 PM

Fluent Tutorial - Laminar Pipe Flow Step #7

( click pict ure for larger image)

This plot shows t hat t he result s did not change by increasing t he fineness of t he mesh.

Thus, we can conclude t hat our 30x100 mesh was good enough. I t is also import ant t o

verify t hat t he solut ion does not change by refining t he mesh in t he st reamwise

direct ion. I n t his case, t he mesh in t he st reamwise direct ion is already fine enough t o

eliminat e mesh- dependent numerical error.

Copyright 2002.

Cornell University

Sibley School of Mechanical and Aerospace Engineering.

Fluent Short Course-Tutorial List | Feedback

http://instruct1.cit.cornell.edu/courses/fluent/plate/step7.htm (7 of 7)11/7/2005 7:09:12 PM

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat

61 2

0.0000000 0.0000000 0

0.0005000 0.0023390 0

0.0010000 0.0037271 0

0.0020000 0.0058025 0

0.0040000 0.0089238 0

0.0080000 0.0137350 0

0.0120000 0.0178581 0

0.0200000 0.0253735 0

0.0300000 0.0330215 0

0.0400000 0.0391283 0

0.0500000 0.0442753 0

0.0600000 0.0487571 0

0.0800000 0.0564308 0

0.1000000 0.0629981 0

0.1200000 0.0686204 0

0.1400000 0.0734360 0

0.1600000 0.0775707 0

0.1800000 0.0810687 0

0.2000000 0.0839202 0

0.2200000 0.0861433 0

0.2400000 0.0878308 0

0.2600000 0.0890840 0

0.2800000 0.0900016 0

0.3000000 0.0906804 0

0.3200000 0.0911857 0

0.3400000 0.0915079 0

0.3600000 0.0916266 0

0.3800000 0.0915212 0

0.4000000 0.0911712 0

0.4200000 0.0905657 0

0.4400000 0.0897175 0

0.4600000 0.0886427 0

0.4800000 0.0873572 0

0.5000000 0.0858772 0

0.5200000 0.0842145 0

0.5400000 0.0823712 0

0.5600000 0.0803480 0

0.5800000 0.0781451 0

0.6000000 0.0757633 0

0.6200000 0.0732055 0

0.6400000 0.0704822 0

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat (1 of 3)12/2/2005 2:52:34 PM

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat

0.6600000 0.0676046 0

0.6800000 0.0645843 0

0.7000000 0.0614329 0

0.7200000 0.0581599 0

0.7400000 0.0547675 0

0.7600000 0.0512565 0

0.7800000 0.0476281 0

0.8000000 0.0438836 0

0.8200000 0.0400245 0

0.8400000 0.0360536 0

0.8600000 0.0319740 0

0.8800000 0.0277891 0

0.9000000 0.0235025 0

0.9200000 0.0191156 0

0.9400000 0.0146239 0

0.9600000 0.0100232 0

0.9700000 0.0076868 0

0.9800000 0.0053335 0

0.9900000 0.0029690 0

1.0000000 0 0

0.0000000 0.0000000 0

0.0005000 -.0046700 0

0.0010000 -.0059418 0

0.0020000 -.0078113 0

0.0040000 -.0105126 0

0.0080000 -.0142862 0

0.0120000 -.0169733 0

0.0200000 -.0202723 0

0.0300000 -.0226056 0

0.0400000 -.0245211 0

0.0500000 -.0260452 0

0.0600000 -.0271277 0

0.0800000 -.0284595 0

0.1000000 -.0293786 0

0.1200000 -.0299633 0

0.1400000 -.0302404 0

0.1600000 -.0302546 0

0.1800000 -.0300490 0

0.2000000 -.0296656 0

0.2200000 -.0291445 0

0.2400000 -.0285181 0

0.2600000 -.0278164 0

0.2800000 -.0270696 0

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat (2 of 3)12/2/2005 2:52:34 PM

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat

0.3000000 -.0263079 0

0.3200000 -.0255565 0

0.3400000 -.0248176 0

0.3600000 -.0240870 0

0.3800000 -.0233606 0

0.4000000 -.0226341 0

0.4200000 -.0219042 0

0.4400000 -.0211708 0

0.4600000 -.0204353 0

0.4800000 -.0196986 0

0.5000000 -.0189619 0

0.5200000 -.0182262 0

0.5400000 -.0174914 0

0.5600000 -.0167572 0

0.5800000 -.0160232 0

0.6000000 -.0152893 0

0.6200000 -.0145551 0

0.6400000 -.0138207 0

0.6600000 -.0130862 0

0.6800000 -.0123515 0

0.7000000 -.0116169 0

0.7200000 -.0108823 0

0.7400000 -.0101478 0

0.7600000 -.0094133 0

0.7800000 -.0086788 0

0.8000000 -.0079443 0

0.8200000 -.0072098 0

0.8400000 -.0064753 0

0.8600000 -.0057408 0

0.8800000 -.0050063 0

0.9000000 -.0042718 0

0.9200000 -.0035373 0

0.9400000 -.0028028 0

0.9600000 -.0020683 0

0.9700000 -.0017011 0

0.9800000 -.0013339 0

0.9900000 -.0009666 0

1.0000000 0 0

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/vertices.dat (3 of 3)12/2/2005 2:52:34 PM

Farfield Boundary - Vertices and Edges

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/01farfield_edges.htm (1 of 2)12/2/2005 2:56:18 PM

Farfield Boundary - Vertices and Edges

http://instruct1.cit.cornell.edu/courses/fluent/airfoil/01farfield_edges.htm (2 of 2)12/2/2005 2:56:18 PM