Tentamen i CFD DEL1

F6L5008 CFD in Design and Development ( del 1 Matematik )

CFD i design och utveckling

Lrare: Armin Halilovic

Examinator: Sture Holmberg
KTH Syd, Campus Haninge

Datum tid: 24/5 frn 8: 15 -11:15

Tilltna hjlpmedel:, boken Computational fluid Dynamics, H K Versteeg, W Malalasekera)
, formelblad (och /eller handboken BETA), minirknare.
Fr godknd krvs 12 p
1) (4p)
Which one (if any) of the following functions
a) 1 ( x, y, z ) = x + y + z
4

2 ( x, y , z ) = x 2 + y 2 + z
2
2
2
c) 3 ( x, y, z ) = x + y + z
b)

satisfies the equation

r
div (U ) = div (grad ) + S

?
r
Here = 5 , U = (1, 2, 3) and S = 2 x + 4 y 23
2) (4p)
r
r
r
r
Find the flux of the vector field F = x i + 2 y j + zk upward through the surface

z = 1+ x2 + 2y , 0 x 1, 0 y 2 .
3) (4p)
r
r
r
r
Use the Divergence Theorem to find the flux of the vector field F = x 2 i + y 2 j + z 2 k out of

the cub:

4) ( 6p)

0 x 1 , 0 y 1 , 0 z 1.

A property is transported by means of convection and diffusion through the domain

sketched in Figure1. The length ( 1m ) is divided into five equal control volumes as shown in
Figure 1. The governing equation is
d d d
= + 20 x 5
dx dx dx
where the dependent variable is .
Boundary conditions:

 is constant at the end points A and B: A = 0 , B = 25 .

Figure 1.

Use the finite volume method and central differencing scheme to determine the
discretised equations for nodal points 1, 2,3,4 and 5.

5) ( 6p)
A plate of thickness 2 cm is shown in Figure 2. The two-dimensional steady state heat transfer
is governed by

T
(k
) + (k
)+S =0
x x
y y

The thermal conductivity of the plate material is k = 500 W/(mC).

The source term is S=4 000 W/m
The north boundary receives a steady heat flux of 200 000 W/m
The south and east boundaries are insulated.
The west boundary is maintained at a temperature of 50 C
Grid: x =0.2m, y= 0.2 m , z=0.02m

Figure 2.

a) Determine the discretised equation at node T11

b) Determine the discretised equation at node T41
c) Determine the discretised equation at node T14

Solutions:
1) (4p)
Which one (if any) of the following functions

a) 1 ( x, y, z ) = x + y + z
4

2 ( x, y , z ) = x 2 + y 2 + z
2
2
2
c) 3 ( x, y, z ) = x + y + z
b)

satisfies the equation

r
div (U ) = div (grad ) + S

?
r
Here = 5 , U = (1, 2, 3) and S = 2 x + 4 y 23
Solution :

r
div(U ) = div(grad ) + S
5 5 5
div( , 2 , 3 ) = div(
,
,
) + 2 x + 4 y 23
x y z
2 3 5 2 5 2 5 2
+
+
=
+
+ 2 + 2 x + 4 y 23
x
y
z
x 2
y 2
z

(eq 1.)

a) Let = 1 ( x, y, z ) = x + y + z
4

Vi calculate the derivatives of 1 and substitute in the left hand side (LHS) and right hand
side of the equation (eq1).
LHS:

RHS=

2 3
+
+
= 4 x3 + 4 y + 3
x
y
z

5 2 5 2 5 2
+
+ 2 + 2 x + 4 y 23 = 60 x 2 + 2 x + 4 y 13
2
2
x
y
z

Whence LHS RHS

Thus the function 1 ( x, y, z ) = x + y + z is not a solution to the equation
4

b) = 2 ( x, y , z ) = x + y + z
LHS= 3 + 2 x + 4 y , RHS= 3 + 2 x + 4 y
2

Whence LHS RHS , and the function

equation

2 ( x, y, z ) = x 2 + y 2 + z is not a solution to the

c) Let = 3 ( x, y, z ) = x + y + z
Then
LHS= 2 x + 4 y + 6 z , RHS= 7 + 2 x + 4 y
2

Thus LHS RHS , and the function 3 ( x, y, z ) = x + y + z

equation.
Answer: None of the functions satisfies the equation
2

is not a solution to the

2) (4p)
r
r
r
r
Find the flux of the vector field F = x i + 2 y j + zk upward through the surface

z = 1+ x2 + 2y , 0 x 1, 0 y 2 .
Solution:
z = 1+ x2 + 2y
r
N = ( z x , z y , 1) = (2 x,2, 1)
r
F = ( x, 2 y , z )
r r
F o N = 2 x 2 4 y + z = 2 x 2 4 y + 1 + x 2 + 2 y = x 2 2 y + 1
1
2
r r
= F o N dxdy = dx ( x 2 2 y + 1)dy =
0

2
3
dx x y y + y
0

]02 = (2 x
1

2)dx =

8
3

Answer: = 8 / 3
3) (4p)
r
r
r
r
Use the Divergence Theorem to find the flux of the vector field F = x 2 i + y 2 j + z 2 k out of

the cub:

0 x 1 , 0 y 1 , 0 z 1.

3) Solution:
r
div( F ) = 2 x + 2 y + 2 z
r
= divFdV = (2 x + 2 y + 2 z )dxdydz =
K

1 1 1

1 1

2
(2 x + 2 y + 2 z )dxdydz = 2 xz + 2 yz + z
0 0 0

0 0

[2 x + 2 y + 1]dxdy = [2 xy + y
1 1

0 0

]10 dxdy =

1
+ y dx =
0

[2 x + 2]dx =3
0

Answer: = 3

4) ( 6p)

A property is transported by means of convection and diffusion through the domain

sketched in Figure1. The length ( 1m ) is divided into five equal control volumes as shown in
Figure 1. The governing equation is
d d d
= + 20 x 5
dx dx dx

where the dependent variable is .

Boundary conditions:
is constant at the end points A and B: A = 0 , B = 25 .
Figure 1.

Use the finite volume method and central differencing scheme to determine the
discretised equations for nodal points 1, 2,3,4 and 5.

Solution:

Integration of the equation

d d d
= + 20 x 5
dx dx dx

(equ 1.1)

over a control volume gives

d d
d
dV = dV + (20 x 5)dV

dx
dx dx
V
V
V
(cancelling A)
e d e
= + (20 P 5)x
w dx w
( (e) ( w) ) = d d + (4 x P 1)
dx e dx w

( )

(equ 1.2)

We use equation 1.2 to get the discretised equations for nodal points 1,2,3, 4 and 5.

Nodes 2,3,4:
Substituting

x
= 0.2 p 0.1, p = 1,2,3,4,5
2
[ x1 = 0.10 , x 2 = 0.30 , x3 = 0.50 , x 4 = 0.70 , x5 = 0.90 ]
+ P
+ P
e = E
, w = W
2
2

d
P
= 5( E P )
= E
x
dx e
and
W
d
= 5( P W )
= P
x
dx w
in the equation
( (e) ( w) ) = d d + (4 x P 1) (equ 1.2)
dx e dx w
x P = a + x p

we get
E + P W + P

= 5( E P ) 5( P W ) + 4 x P 1
2
2
10 P = 5.5W + 4,5 E + 4 x P 1
Substituting P=2, 3, 4 we obtain
10 2 = 5.51 + 4,53 + 0.2
103 = 5.5 2 + 4,5 4 + 1
10 4 = 5.53 + 4,55 + 1.8

( equ P2)
(equ P3)
(equ P4)

Node 1

Substituting e =

E + P
2

w = A

, x1 = 0.10

P
d
= 5( E P ) and
= E
x
dx e
A
d
= 10( P 0) = 10 P
= P
x / 2
dx w
into
( (e) ( w) ) = d d + (4 x P 1)
dx e dx w
we obtain

(equ 1.2)

E + P

A = 5( E P ) 10 P 0.6
2
15.5 P = 4,5 E 0.6
which, for p=1, implies
15.51 = 4,5 2 0.6

Node 5

Substituting w =

W + P
,
2

E = B = 25

, x5 = 0.90 ,

W
d
= 5( P W ) and
= P
x
dx w

d B P
= 10(25 P ) = 10P 250
=
e

/
2
dx
x

into

( (e) ( w) ) = d

d
+ (4 x P 1)
dx e dx w

(equ 1.2)

we obtain
+ P
B W
= (250 10 P ) 5( P W ) + 2.6
2
14.5 P = 5,5W = 227.6
which, for p=5, implies
14.5P = 5,5W = 227.6 (equ P5)
5. Answer:

The discretised equations for nodal points 1,2,3, 4 and 5 are:

15.51 = 4,5 2 0.6
( equ P1)
10 2 = 5.51 + 4,53 + 0.2
( equ P2)
103 = 5.5 2 + 4,5 4 + 1
(equ P3)
(equ P4)
10 4 = 5.53 + 4,55 + 1.8
14.5P = 5,5W = 227.6
(equ P5)
Commentary:
This was not a part of the question, but if we solve the system then we get
the numerical solution:

[ 1.50, 5.30, 9.90, 15.3, 21.5 ]

We can compare numerical with the exact solution
= 10 x 2 + 15 x

5) ( 6p)
A plate of thickness 2 cm is shown in Figure 2. The two-dimensional steady state heat transfer
is governed by

T
(k
) + (k
)+S =0
x x
y y
The thermal conductivity of the plate material is k = 500 W/(mC).
The source term is S=4 000 W/m
The north boundary receives a steady heat flux of 200 000 W/m
The south and east boundaries are insulated.
The west boundary is maintained at a temperature of 50 C
Grid: x =0.2m, y= 0.2 m , z=0.02m

Figure 2.

a) Determine the discretised equation at node T11

b) Determine the discretised equation at node T41
c) Determine the discretised equation at node T14

Solution:

Integration of the equation

T
(k
) + (k
)+S =0
x x
y y

(equ 1.1)

over a control volume gives

d dT
d dT
dV + SdV = 0
dV +

dx
dx
dy
dy

V
V
V
T n
T e
+ Sxyz = 0
Ayz k
+ Axz k

y
x w
s

T
T
T
T
+ Sxyz = 0
As k
Ae k
+ An k
Aw k
n

y
y
x w
x e
s

Substitution of numerical values Sxyz = 8 gives

T
T
T
T
+ 3.2 = 0 (equ 2.1)
As k
Ae k
+ An k
Aw k
x w
x e
y s
y n
[ (FluxE-FluxW) + (FluxN-FluxS)+3.2=0 ]
We use equation (equ 2.1) to derive the discretised equations for nodal points
T15 and T41.
The constant values in the equation are:
Ae = Aw = yz = 0.004 ,
An = As = xz = 0.004 , and
k=500,

Method 1)
Point T11 :
At nodal point T 11 we have

T TB
T TB
T
FluxW = Aw k
=2 P
= 20(TP 50) = 20TP 1000
=2 P
h/2
0.1
x w
T TP
T TP
T
FluxE = Ae k
=2 E
= 10(TE TP )
=2 E
h
0.2
x e
T TP
T
= 2 N
FluxN = An k
= 10(TN TP )
h
y n
FluxS = 0
Substitution in equation (equ 2.1) gives
10(TE TP ) 20(TP 50) + 10(TN TP ) + 3.2 = 0
40TP = 10TE + 10TN + 1003.2
40T11 = 10T12 + 10T21 + 1003.2 (equT11)
Point T14 :
At nodal point T 11 we have

T TB
T TB
T
FluxW = Aw k
=2 P
= 20(TP 50) = 20TP 1000
=2 P
h/2
0.1
x w
T TP
T TP
T
FluxE = Ae k
=2 E
= 10(TE TP )
=2 E
h
0.2
x e
FluxN = An 200000 = 0.004 200000 = 800
T TS
T TS
T
= 2 P
FluxS = As k
=2 P
= 10(TP TS )
h
0.2
y s
Substitution in equation (equ 2.1) gives
10(TE TP ) 20(TP 50) + 800 10(TP TS ) + 3.2 = 0
40TP = 10TE + 10TS + 1803.2
40T14 = 10T24 + 10T13 + 1003.2
(equT14)

Point T41 :
At nodal point T 41 we have

T TW
T TW
T
FluxW = Aw k
=2 P
= 10(TP TW )
=2 P
h
0.2
x w
FluxE = 0
T TP
T TP
T
= 2 N
FluxN = An k
=2 N
= 10(TN TP )
h
0.2
y n
FluxS = 0

Substitution in equation (equ 2.1) gives

0 10(TP TW ) + 10(TN TP ) 0 + 3.2 = 0
20TP = 10TW + 10TN + 3.2
20T41 = 10T31 + 10T42 + 3.2 (equT41)
Answer:
40T11 = 10T12 + 10T21 + 1003.2
(equT11)
(equT14)
40T14 = 10T24 + 10T13 + 1003.2
20T41 = 10T31 + 10T42 + 3.2
(equT41)

Method 2)
We start with equation
T
T
T
T
+ 3.2 = 0
As k
Ae k
+ An k
Aw k
x w
x e
y s
y n
The constant values in the equation are:
Ae = Aw = yz = 0.004 ,
An = As = xz = 0.004 , and
k=500.

(equ 2.1)

We calculate the coefficients a P , aW , a E , a N , a S , S P and S u

k Aw 500 0.004
k Ae
aE =
=
= 10 ,
= 10
x
0 .2
x
k As 500 0.004
k An
=
= 10 , a N =
= 10
aS =
y
0.2
y
aW =

S u ( source) = 3.2 ,

S p ( source) = 0

Boundary node T11

aW = 0 , a E = 10 ,
a S = 0 , a N = 10
2 k Aw 50
+ 0 = 3.2 + 100 = 1032 (Fixed value at the west
x
boundary and insulated at the south boundary)
2 k Aw
S p (total)= S p ( source) + (
) + 0 = 20
x

S u (total)= S u ( source) +

Boundary node T14 ((Fixed value at the west boundary and fixed flux at the north
boundary)

aW = 0 , a E = 10 ,

a S = 10 ,

aN = 0

S u (total ) = S u ( source) +

2 * k Aw
TB + An 200000
x

2 500 0.004
50 + 800 = 3.2 + 1000 + 800 = 1803.2
0.2
2 k Aw
S p (total)= S p ( source) + (
) + 0 = 20
x
Boundary node T41 ( The south and east boundaries are insulated.)
= 3.2 +

aW = 10 , a E = 0 ,
a S = 0 , a N = 10
S u (total ) = S u ( source) + 0 + 0 = 3.2
S p (total)= S p ( source) + 0 + 0 = 0

Node
T11
T14
T41

aW
0
0
10

aE
10
10
0

aS
0
10
0

aN
10
0
10

S P (total)

-20
-20
0

S u (total)

a P = aW + aE + a N + a S

1003.2
1800
3.2

- S P (total)
40
40
20

The discretised equation for node P is

a P TP = aW TW + a E TE + a S TS + a S TS + S u (total ) ,
and the equations (for nodal points T15 and T41) are
40T11 = 10T12 + 10T21 + 1003.2
40T14 = 10T24 + 10T13 + 1003.2

20T41 = 10T31 + 10T42 + 3.2

(equT11)
(equT14)
(equT41)