Solution to two-dimensional Incompressible Navier-Stokes Equations withSIMPLE, SIMPLER and Vorticity-Stream Function Approaches.Driven-Lid Cavity Problem: Solution and Visualization.

Maciej Matyka

Computational Physics Section of Theoretical PhysicsUniversity of Wroclaw in PolandDepartment of Physics and AstronomyExchange Student at University of Link¨oping in Swedenmaq@panoramix.ift.uni.wroc.plhttp://panoramix.ift.uni.wroc.pl/

∼

maqMay 8, 2003

Abstract

In that report solution to incompressible Navier - Stokes equations in non - dimensional form will be presented.Standard fundamental methods: SIMPLE, SIMPLER (SIMPLE Revised) and Vorticity-Stream function approach arecompared and results of them are analyzed for standard CFD test case - Drived Cavity ﬂow. Diﬀerent aspect ratiosof cavity and diﬀerent Reynolds numbers are studied.

1 Introduction

The main problem is to solve two-dimensional Navier-Stokes equations. I will consider two diﬀerent mathemat-ical formulations of that problem:

•

u,v,p

primitive variables formulation

•

ζ,ψ

vorticity-stream function approachI will provide full solution with both of these methods.First we will consider three standard, primitive componentformulations, where fundamental Navier-Stokes equationwill be solved on rectangular, staggered grid. Then, solu-tion on non-staggered grid with vorticity-stream functionform of NS equations will be shown.

2 Math background

We will consider two-dimensional Navier-Stokes equationsin non-dimensional form

1

:

1

We consider ﬂow without external forces i.e. without gravity.

∂

−→

u∂t

=

−

(

−→

u

)

−→

u

−

ϕ

+1

Re

2

−→

u

(1)

D

=

−→

u

= 0 (2)Where equation (2) is a continuity equation which hasto be true for the ﬁnal result.

3 Primitive variables formulation

First we will examine SIMPLE algorithm which is basedon primitive variables formulation of NS equations. Whenwe say ”primitive variables” we mean

u,v,p

where

u

=(

u,v

) is a velocity vector, and

p

is pressure. We can rewriteequation (1) in diﬀerential form for both velocity compo-nents:

∂u∂t

=

−

∂u

2

∂x

+

∂uv∂y

−

∂p∂x

+1

Re

(

∂

2

u∂x

2

+

∂

2

v∂y

2

) (3)

∂v∂t

=

−

∂v

2

∂y

+

∂uv∂x

−

∂p∂y

+1

Re

(

∂

2

u∂x

2

+

∂

2

v∂y

2

) (4)We rewrite continuity equation in the following form:

∂u∂x

+

∂v∂y

= 0 (5)These equations are to be solved with SIMPLE method.

3.1 SIMPLE algorithm

SIMPLE algorithm is one of the fundamental algorithm tosolve incompressible NS equations. SIMPLE means: SemiImplicit Method for Pressure Linked Equations.Algorithm used in my calculations is presented in theﬁgure (1). First we have to guess initial values of thepressure ﬁeld

2

(

P

∗

)

n

and set initial value of velocity ﬁeld- (

U

∗

)

n

, (

V

∗

)

n

. Then equation (3) and (4) is solved toobtain values of (

U

∗

)

n

+1

, (

V

∗

)

n

+1

. Next we have to solvepressure-correction equation:

2

p

=1∆

t

(

·

V

) (6)After that we use simple relations to obtain correctedvalues of pressure and velocity ﬁelds. Then we can checkif solution is coverged.

2

Subscripts denote computational step, where ”n+1” means cur-rent step.

Guess(P*)

n

, (U*)

n

, (V*)

n

Solve Momentumeq. for(U*)

n+1

, (V*)

n+1

Solve Poisson eq.for (P’)P

n+1

= (P*)

n

+ (P’)U

n+1

= (U*)

n

+ (U’)V

n+1

= (V*)

n

+ (V’)

If not covergedIf coverged

Visualize Results

Figure 1: SIMPLE Flow Diagram

3.2 Numerical Methods in SIMPLE

3.2.1 Staggered Grid

For discretization of diﬀerential equations I am using stag-gered grid. In the ﬁgure (2) staggered grid for rectangulararea is sketched. Primitive variables are placed in diﬀerentplaces. In points

i,j

on a grid pressure

P

values, in points

i

+0

.

5

,j u

x-velocity components and in points

i,j

+0

.

5

v

y-velocity components are placed. That simple model of staggered grid gives us possibility to use simple discretiza-tion with second order accuracy which will be discussedlater.

u

i+0.5,j

p

i,j

v

i,j-0.5

v

i,j+0.5

u

i-0.5,j

p

0,0

u

0.5,0

v

0,0.5

Figure 2: Staggered grid: ﬁlled circles

P

, outline circles

U

x-velocity, cross

V

y-velocity component.2

Diﬀerential Discretization Type

∂u∂tu

n

+1

−

u

n

∆

t

forward,

O

(

h

)

∂

2

u∂x

2

u

i

+1

,j

−

2

∗

u

i,j

+

u

i

−

1

,j

(∆

x

)

2

central,

O

(

h

2

)

∂u

2

∂xu

2

i

+1

,j

−

u

2

i

−

1

,j

(2

·

∆

x

)

central,

O

(

h

2

)

∂p∂x p

i

+1

,j

−

p

i,j

(∆

x

)

forward,

O

(

h

)

∂p∂y p

i,j

+1

−

p

i,j

(∆

x

)

forward,

O

(

h

)Table 1: Discretizations used in SIMPLE algorithm

3.2.2 Discretization Schemes

Let us now examine some numerical methods used in pre-sented solution. For algorithm presented in the ﬁgure (1)we have only three equations which have to be discretizedon a grid. First we have momentum equations (3) and (4).Discrete schemes used in discretization of momentumequations are presented in a table (1). Using presenteddiscrete form of derivatives I obtain numerical scheme formomentum equations exactly in the form presented in [1].Equations (3) , (4) discretized on staggered grid can bewritten

3

as follows

4

:

u

n

+1

i

+0

.

5

,j

=

u

ni

+0

.

5

,j

+ ∆

t

·

(

A

−

(∆

x

)

−

1

(

p

i

+1

,j

−

p

i,j

)) (7)

v

n

+1

i,j

+0

.

5

=

u

ni,j

+0

.

5

+ ∆

t

·

(

B

−

(∆

y

)

−

1

(

p

i,j

+1

−

p

i,j

)) (8)where

A

and

B

are deﬁned as:

A

=

−

a

1

+ (

Re

)

−

1

·

(

a

3

+

a

4

) (9)

B

=

−

b

1

+ (

Re

)

−

1

·

(

b

3

+

b

4

) (10)and respectively we deﬁne:

3

Please note than cited [1] reference contains some print mistakesthere.

4

Generally I show there only an idea how to write discretizedequations, they should be rewritten with ”*” and ”’” chars for con-crete steps of the algorithm

a

1

=

−

(

u

2

)

ni

+1

.

5

,j

−

(

u

2

)

ni

−

0

.

5

,j

2

·

∆

x

−

(

u

˙

v

)

ni

+0

.

5

,j

+1

−

(

u

˙˙

v

)

ni

+0

.

5

,j

−

1

2

·

∆

y

(11)

b

1

=

−

(

v

2

)

ni,j

+1

.

5

−

(

v

2

)

ni,j

−

0

.

5

2

·

∆

y

−

(

v

˙

u

)

ni

+1

,j

+0

.

5

−

(

v

˙˙

u

)

ni

−

1

,j

+0

.

5

2

·

∆

x

(12)(

a

3

) =

u

ni

+1

.

5

,j

−

2

·

u

ni

+0

.

5

,j

+

u

ni

−

0

.

5

,j

(∆

x

)

2

(13)(

a

4

) =

u

ni

+0

.

5

,j

+1

−

2

·

u

ni

+0

.

5

,j

+

u

ni

+0

.

5

,j

−

1

(∆

y

)

2

(14)(

b

3

) =

v

ni,j

+1

.

5

−

2

·

v

ni,j

+0

.

5

+

v

ni,j

−

0

.

5

(∆

y

)

2

(15)(

b

4

) =

v

ni

+1

,j

+0

.

5

−

2

·

v

ni,j

+0

.

5

+

v

ni

−

1

,j

+0

.

5

(∆

x

)

2

(16)Now we have deﬁned almost everything. Dotted veloc-ities should be also deﬁned. I use simple expressions forit:˙

u

= 0

.

5

·

(

u

i

−

0

.

5

,j

+

u

i

−

0

.

5

,j

+1

) (17)˙˙

u

= 0

.

5

·

(

u

i

+0

.

5

,j

+

u

i

+0

.

5

,j

+1

) (18)˙

v

= 0

.

5

·

(

v

i,j

+0

.

5

+

v

i

+1

,j

+0

.

5

) (19)˙˙

v

= 0

.

5

·

(

v

i,j

−

0

.

5

+

v

i

+1

,j

−

0

.

5

) (20)

3.2.3 Poisson Equation

For equation I use simple iterative procedure. In the ﬁgure(3) points used for calculation of pressure at each (

i,j

) gridpoints are marked.

P

i,j

P

i+1,j

P

i-1,j

P

i,j-1

P

i,j+1

Figure 3: Points on a grid used in iterative procedure forPoisson equation solving.3

Filters