International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

doi: 10.14355/ijnese.2013.0304.02

www.ijnese.org

Study on Influence of Rewetting on

Conduction Heat Transfer for AHWR Fuel
Bundle Re-flooding Phenomena
Mithilesh Kumar1, D. Mukhopadhay1, A. K. Ghosh2, Ravi Kumar3
Reactor Safety Division1, Bhabha Atomic Research Centre, Mumbai, India
Raja Ramanna Fellow2, Bhabha Atomic Research Centre, Mumbai, India
Indian Institute of Technology, Roorkee3, India
mithiles@barc.gov.in

Abstract
Numerical study on re-flooding behavior of AHWR fuel
bundle has been carried out to assess influence of
circumferential and axial rewetting on the conduction heat
transfer. As the AHWR fuel bundle quenching under
accident condition is designed primarily with radial jets at
several axial locations, however, bottom re-flooding still
remain as an another option. A 3-D (r, , z) transient
conduction fuel pin model has been developed to carry out
the study with a Finite Difference Method (FDM) technique
with Alternating Direction Implicit (ADI) scheme. Both the
analyses are carried out with same fluid temperature and
heat transfer coefficients as boundary conditions. It has been
found from the analyses that for radial jet, the
circumferential conduction is significant and overall the fuel
temperature falls in the quench plane with the initiation of
quenching event. Axial conduction is found to be very poor
for bottom reflooding case and the fuel temperature only
falls when the quench front reaches the location. Sensitivity
studies with respect to the direction of solving by ADI are
carried out to assess directional influence on the solution.
The study shows that direction (r- -z) of solution is
insensitive to the solution for axial and circumferential
conduction solutions.
Keywords
Re-flooding; Rewetting; Heat Conduction; Numerical Method;
Sensitivity Study

Introduction
Rewetting of hot surface is a process in which a liquid
wets a hot surface by displacing its own vapour that
otherwise prevents the contact between the solid and
liquid phases. This has generated immense interest in
studying rewetting through both theoretical
simulation by Yamanouchi, Coney M.W.E. and
experimental studied was carried by Yamanouchi,
Duffey R.B. Falling film rewetting for several vertical
geometries such as plates (Coney M.W.E., Tien C.L.),

rods [Blair J.M.] and tubes [Satapathy A.K.] have been

modeled by a number of researchers. In general, in all
models, a moving rewetting front that divides the
solid into two distinct regions is considered. Most of
the models also consider a constant rewetting velocity
that reduces the problem into a quasi-static one. Initial
efforts were made to formulate one-dimensional
conduction models [Yamanouchi] that are reasonably
successful in correlating rewetting phenomena at low
Peclet number. Tien and Yao presented the asymptotic
solutions of a two-dimensional conduction model
which clearly demonstrates the different physical
pictures for the cases of high and low coolant flow
rates. A variety of techniques have been used for
solving two-dimensional conduction models for falling
film rewetting. Some of the important studies are
elaborated. Because of mathematical difficulty, most
two-dimensional analyses are either approximate or
numerical ones. Duffey and Porthouse first considered
for solving the rewetting problem by separation of
variables. They retained only the first term in the series
solution. However, Coney M.W.E reported that using
a small number of terms in a series yields inaccurate
results due to slow convergence of the series. An
approximate solution to the same model for a
cylindrical rod was presented by Blair J.M. Tien and
Yao first applied the WienerHopf technique to a twodimensional rewetting problem of a rectangular slab,
while an exact solution to the same problem was
presented by Castiglia et al., employing the method of
separation of variables. Numerical solutions of
conduction controlled rewetting were provided by
Satapathy et al., Thompson, and Raj and Date by using
the finite difference technique. Heat Balance Integral
Method (HBIM) is one of many semi-analytical
methods used to solve conduction problems (Eckert
and Goodman). This is analogous to classical integral

85

www.ijnese.org

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

technique used for fluid flow and convective heat

transfer analysis. A numerical study has been made to
investigate the effect of internal heating and
precursory cooling during quenching of an infinite
tube was studied by Satapathy A. K (2000).

location. However, the developed computational

model is equally applicable for 2nd and 3rd circle pins
with a different boundary conditions.

An experimental assessment of rewetting of a

Advanced Heavy Water Reactor (AHWR) specific
nuclear fuel bundle with radial jets has been carried
out by Patil N. D (2013). The experimentation involved
cooling of a 54-rod bundle by in-bundle injection and
demonstrated that quenching occurs for all fuel pins
along the radial direction. However, the study did not
bring out the radial jet influence on circumferential
rewetting of a fuel pin.
The numerical study presented in this paper aimed to
bring out a comparison in rewetting pattern for a
radial jet rewetting versus axial rewetting for AHWR
fuel bundle. The proposed reactor is a vertical pressure
tube type, heavy water moderated boiling light water
coolant reactor. Fuel bundle is housed in the pressure
tube (PT) which in turn is housed in calandria tube
(CT). The 3.5 m height fuel bundle has 54 fuel pins
arranged in three different concentric rings as shown
in Fig.1. The inner, middle and outer rings having 12,
18 and 24 fuel pins and centre of the fuel pin have a
water rod which is used to inject emergency coolant in
radial direction at different elevations of the fuel
bundle under a pipe break scenario. The water rod has
eight holes of 1.5 mm dia. at one elevation along the
circumference of water rod. The radial injection is
designed for 13 different axial locations along the 3.5
m length of water rod. More details of AHWR fuel
bundle and its injection arrangement have been
furnished by Sinha et al. Bottom re-flooding is also an
option with the designer where water may be injected
from bottom most holes instead on 13 axial elevations.
Under this study, a 3-D (r, , z) transient conduction
fuel pin model has been developed based on Finite
Difference Method (FDM) technique with Alternating
Direction Implicit (ADI) scheme. The number of nodes
in r, & z direction has been considered to be 50, 20
and 50 respectively. A 1.6 ms time step has been used
as predicted from stability criteria. A total time of 4 hrs
is required for individual rewetting study.
Single pin from first circle has been considered for
which bottom and radial reflooding was studied. The
1st circle pin is selected as it experiences a strong radial
jet and experience a large circumferential temperature
gradient. For bottom rewetting case, all the pins will
experience the same velocity front irrespective of its
86

FIG. 1 VIEW OF AHWR FUEL BUNDLE

Both the analyses are carried out with same fluid

temperature and heat transfer coefficients as boundary
conditions. The single fuel pin considers for
circumferential and axial rewetting. The three
dimensional partial differential equation for unsteady
state conduction equation for cylindrical rod is as
follows,
2T 1 T 1 2T 2T Q 1 T
(1)
+
+
+
+
=
K t
r 2 r r r 2 2 z 2
Alternating Direction Implicit (ADI) numerical method
has been adopted. With an ADI method, the heat
diffusion equation is first solved implicitly in the rdirection while leaving the other two directions
explicit. The heat diffusion equation is then solved
implicitly in a similar way in the and z direction.
This scheme reduces a three-dimensional problem to a
series of one-dimensional implicit problem. Fig 2
shows the three different steps in ADI.
Many methods [Peaceman D.W, Brian, Douglas J.,
Chang, M] are also available in ADI for solving the
heat diffusion equation, as Peaceman-Rachford pureADI method, Brain method, Douglas method, and
based on superposition principle by Jules Thibault
(1994). All above schemes having the same problem of
conditionally stable criteria as right side of equations
have a negative coefficient. This paper discussed
conduction effect on axial, circumferential direction as
well as the study on sensitivity on the direction of
solution(r- - z) in steady state as well as in transient
which is not found in open literature. This study is
required to assess the time frame to reach the steady
state condition.

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

n +1/3
n +1/3
Ti n+1,+1/3
j , k 2Ti , j , k + Ti 1, j , k

Ti ,nj +1, k

r 2
2Ti ,nj , k + Ti ,nj 1, k

ri2 2
n +1/3
n
Q
1 Ti , j , k Ti , j , k
+
=
k
t / 3
f

www.ijnese.org

n +1/3
n +1/3
1 Ti +1, j.k Ti 1, j ,k
+
2r
ri
Ti ,nj , k +1 2Ti ,nj , k + Ti ,nj , k 1

z 2

(1.1)

Or

r n +1/3
T
= F0 t 1
+
2r i 1, j , k
i

r n +1/3 2 F t
F0 t 1 +
T
+ 1
2 Fz t Ti n, j , k +
2
2r i +1, j , k

ri
i

F t n
Q t
n
n
n
Ti , j +1, k + Ti , j 1, k + Fz t Ti , j , k +1 + Ti , j , k 1 +
2
ri
3 f C p

(1 + 2F t ) T

n +1/3
i, j ,k

FIG. 2 ADI STEPS TO SOLVE THE CONDUCTION EQUATION

The tri-diagonal matrix has been used to solve the

equation corresponding to given direction. Normally
in ADI scheme, first solution is obtained in r-direction
in an implicit manner and leaves both directions
explicit. But here all the possible six combinations,
likes r- - z, r- z- , - r- z, - z- r, z- r- and z - - r
have been used to solve the heat diffusion equation
and compared results.
3-D Model Development
Finite Difference Formulation
FDM is used to formulate the heat diffusion equation
in cylindrical Co-ordinate (Eq. no. 1) and further semi
implicit used to formulate in r--z direction. The
differential equation has been divided into the three
category (1) centre of fuel pin (2) between centre of
fuel pin and surface and (3) surface of fuel pin. The

length increment in r, , z directions are r , z

respectively. The semi implicit scheme has been used
to formulate the differential equation in r, and z
directions and is represented by equation numbers 1.1,
1.2, and 1.3 respectfully. As per ADI norms, first 1/3
time increment has been taken to solve the
temperature at different grid points in r-direction next
1/3 time increment for solving the temperature in
direction and last time step, for solving the
temperature in z-direction. For stability criteria, the
coefficient of Ti ,nj+.k1/3 , Ti ,nj+.k2/3 , Ti ,nj+.k1 in right side of the
equation must be positive.
Finite difference formulation between centre and
surface of fuel rod
r-direction

Condition for stability

2 F t

2 Fz t 0
1
2

ri

-direction
n +1/3
n +1/3
Ti n+1,+1/3
j , k 2Ti , j , k + Ti 1, j , k

Ti n, j++2/3
1, k

r 2
2Ti n, j+, k2/3 + Ti n, j+2/3
1, k

+
ri2 2
n + 2/3
n +1/ 3
Q
1 Ti , j , k Ti , j , k
+
=
t / 3
k
f

n +1/3
n +1/3
1 Ti +1, j , k Ti 1, j ,k
+
2r
ri
n +1/3
n +1/3
Ti n, j+,1/3
k +1 2Ti , j , k + Ti , j , k 1

z 2

(1.2)

Or
2 F t n + 2/3 F t n + 2/3
Ti , j +1, k + Ti n, j+2/3
1 +
Ti , =
j ,k
1, k +
2
2
r
r
i
i

r n +1/3
T
1 2 F0 t 2 Fz t Ti n, j+,1/3
1+
+
k + F0 t
2r i +1, j , k
i

Q t
r n +1/3
n +1/3
F0 t 1
T
+ Fz t Ti n, j+,1/3
k +1 + Ti , j , k 1 +
2r i 1, j , k
3 f C p
i

Condition for stability

(1 2F t 2F t ) 0
0

z- direction
n + 2/3
n + 2/3
Ti n+1,+ 2/3
j , k 2Ti , j , k + Ti 1, j , k

Ti n, j++2/3
1, k

r 2
2Ti n, j+, k2/3 + Ti n, j+2/3
1, k

ri2 2
n +1
n + 2/3
1 Ti , j , k Ti , j , k
Q
+
=
k
t / 3
f

n + 2/3
n + 2/3
1 Ti +1, j , k Ti 1, j , k
2r
ri
Ti n, j+,1k +1 2Ti n, j+,1k + Ti n, j+,1k 1
+
(1.3)
z 2

or
Z-direction

Ti ,nj+,1k 1 + 2 Fz t = Fz t Ti ,nj+,1k +1 + Ti ,nj+,1k 1 +

r n + 2/3
r n + 2/3
F0 t 1 +
Ti +1, j , k + F0 t 1
T

2r
2r i 1, j ,k
i
i

87

www.ijnese.org

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

2 F t n + 2/3 F t n + 2/3
n + 2/3
1 2 F0 t
Ti , j , k + 2 Ti , j +1, k + Ti , j 1, k
2
r
r
i
i

Qt
+
3 f C p

Condition for stability

2 F t
1 2 F0 t
0
ri2

Centre and Surface of the Fuel Pin

The centre line temperature for fuel pin has been
calculated by using the concept of energy balance and
finite difference formulation is used to calculate the
centre line temperature. It is shown by Equation 1.4.
The centre line grid point represented by i and next to
centre line temperature is represented by i+1 and is
shown in Fig. 3. In similar way, surface temperature of
fuel pin has been formulated in Equation 1.5.
n
n
r
r Ti , j , k Ti +1

Q 2 k s

r
2
2

1
+
n
n
2

T
T

r
i, j ,k
i, j ,k

= f cp

t
2

(1.4)

or

=
Ti ,nj+,1k 4 F0Ti n+1, j +1, k +1 + Ti n, j , k 1 4 Fo +

Q t

f Cp

Condition for stability

(1 4F ) 0
o

FIG. 3 DISCRITISATION FOR CENTRE LINE TEMPERATURE

At surface of fuel pin

Implicit Scheme

2 R f K f Tin1+1/3 Tin +1/3

) + 2 R h

f o

(T

inf

Ti

n +1/3

n +1/3
2

Tin (1.5)

r
r Ti

Q f C p R 2f R f
+ R 2f R f =

2
2

3
Or

8 Fo R f t
8 Fo R f ho r t n +1/3
1.0 +
T
+
i

K f 4 R f r
4 R f r

8 Fo R f t n +1 Q t 8 Fo R f ho r t
Ti 1 +
Tinf + Ti n
=
+
3K f
K f 4 R f r
4 R f r
2

88

Boundary Condition and Solver

Once the formulation of differential equations is over
then with help of boundary, it is required to calculate
the temperature at different nodes of fuel pin. There
are two cases considered for validation of numerical
code; the first of which is constant coolant temperature
and heat transfer coefficient while second case is
coolant temperature varying along the length of fuel
pin when it flows along the length of pin. The
governing equation which calculates the coolant
average temperature between the corresponding nodes
of fuel pin is represented by Eq.1.7.
Energy balance at surface of fuel pin
ho A(Ts Tinf
=
) mC p (Tout Tin )

(1.6)

Tin + Tout
2
h A
ho ATs
h A
Tout ( o =
+ 1)
+ (1 o )Tin
2mC p
2mC p
mC p
Tinf =

(1.7)
(1.8)

Equation no 1.8 is used for calculating the outlet

coolant temperature for a given volume. After
applying the boundary condition, it is required to
solve the discretised partial differential in each
direction by ADI method. At each incremental time
step, the differential equations form a tri-diagonal
matrix is solved by Thomas Algorithm (Eq 1.9). In first
time increment (i.e. 1/3 s), it solves temperature in rdirection at different grids points and takes old values
of temperature in , z direction. Similarly, for higher
time step (2/3 s), this algorithm solves in direction
and remaining directions take previous value.

:
:

T1,nj+,1/3

k
n +1/3
T2, j , k
T3,n +j ,1/3
k
n +1/3
T4, j , k =
.

Tnn++1,1/3j , k

b1
b
2
b3
b4
.
.
.

bn +1

(1.9)

Different heat transfer correlations have been used to

determine the heat transfer coefficient at the surface of
fuel pin. Thom correlation (Eq. 2.0) is used for nucleate
boiling and Modified Bromley correlation (Eq. 2.1) is
used for film boiling.
p 6.2
e
=
ho
0.79

3
Tsat

k g h fg g g ( Lg )

ho = 0.62
g LTsat

(2.0)
0.25

(2.1)

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

g
c

L = 2
g ( Lg )

Steady State Computation for Uniform and Non-uniform

Heat Generation
Fig 4 shows the 3D temperature behavior in the fuel
pin in different cases, the first of which is uniform heat
generation (1.16E+8 W/m3) with a constant heat
transfer coefficient and coolant temperature and
second case is with sinusoidal heat generation with
variation of coolant temperature along the length of
the fuel pin.

direction. Here sensitivity studies against the direction

of solution like r- - z, -r- z and z- r- have been
carried out. In -r- z, ADI first solves the temperature
in direction implicitly and then goes to r and z
direction and similar ways for left last cases. It has
been found that there is insensitivity of temperature
prediction by ADI method by changing the directions
of solution.
780

760

Temperature (K)

Where

www.ijnese.org

740

r_theta_Z
Theta_r_Z
Z_r_theta

720

700

680

660
0

20

40

60

80

100

Time (s)
Steady state temperature of surface of fuel pin

FIG. 5 STEADY STATE SURFACE TEMPERATURE DISTRIBUTION

BY SENSITIVITY STUDIES IN ADI

380

Temperature in fuel pin has been predicted uniform

across the fuel pin and at any cross section minimum
and maximum temperatures are at the surface and
centre of fuel pin respectively.
For sinusoidal heat generation, as shown in Fig. 4, the
fuel surface temperature starts increasing along the
length of the fuel pin but in half of the fuel pin
temperature first increases very fast duo to increasing
the heat flux, later, it is not decreased even decreases
of flux due to integration of heat energy in the fuel pin
which results in maximum fuel pin surface
temperature in-between second half of the fuel pin and
its end.
Sensitivity Study on Direction of Solution
Sensitivity steady has been carried for calculation of
temperature and fuel surface temperature is shown in
Fig. 5. As practice in open literature, ADI first solves in
r- direction implicitly for calculation of temperature
then goes to -direction and finally it solves in z

360

Temperature (oC)

FIG. 4: TEMPERATURE DISTRIBUTION ON SOLID FUEL PIN IN

(A) UNIFORM (B) NON UNIFORM (SINUSOIDAL) HEAT
GENERATION

Tmax = 380 K at 3.09 m

340
320

25 Nodes
50 Nodes
75 Nodes

300
280
0.0

0.5

1.0

1.5

2.0

2.5

3.0

Length (m)
FIG. 6 STEADY STATE SURFACE TEMPERATURE DISTRIBUTION ALONG THE LENGTH BY SENSITIVITY STUDIES IN
NUMBER OF NODES

Further sensitivity studies have been carried out with

variation of numbers nodes along the length of furl pin.
It has been found that ADI method is insensitivity and
shown in Fig. 6.
Benchmarking of 3-D Model
A benchmarking exercise has been carried out for the
3-D conduction model with the available 1-D
analytical solution for a steady state case. As analytical
solution of 3-D is not available, the benchmarking is

89

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

carried out with 1-D solution. The analytical solution

for estimating the coolant temperature and fuel pin
surface temperature along the length of fuel pin is
represented by equation no. 2.2 and 2.3 [Glasstone S et
al] and used to calculate the fluid temperature and
surface temperature for 1-D analytical solution. The
heat flux along the fuel pin has been considered as a
cosine profile. The Fourier conduction equation and its
analytical solution for calculating the temperatures
along the radius of fuel pin solution are given in
Equation 2.4 and 2.5 respectively. Table 1 provides the
input parameters for this exercise.
Tinf =
Ts
=

Qmax A L

mc p

(1 cos(

x
l

370
360
340
330
320

Coolant temperature

310
300
290

=
0
+
Kf

250
0.0

0.5

1.0

1.5

2.0

2.5

3.0

Time (s)

(2.4)

FIG. 7 BENCHMARKING OF COOLANT AND SURFACE

TEMPERATURE PREDICTION

Fig. 8 shows the comparison of temperature

distribution in r-direction with 3-D numerical model
and analytical solution for fuel rod. Result shows a
good agreement between the two solutions.

At outer surface

950

TABLE 1 INPUT DATA FOR BENCHMARK EXERCISE:

Value
0.0041 m
3.5 m
2.8 W/m K
10730 kg/m3
0.28 KJ/kg K
300 oC
500 W/m2K
x
8
l

Tmax = 938oC

850

Analytical
Numerical

800

750

0.000

0.001

Fig 7 shows the coolant temperature and fuel surface

temperature variation along the length of the fuel pin.
The calculated coolant temperature and surface
temperature using equation 1.7 and 1.8, acts as a

0.003

0.004

FIG. 8 STEADY STATE TEMPERATURE DISTRIBUTION ALONG

THE CENTRE OF FUEL PIN
200

) (W/m3)

Two step benchmarking exercise has been done. In the

first step, boundary conditions used for calculating the
fuel radial temperature distribution for 3-D model and
analytical solution are compared and in the
subsequent step the radial temperature distributions
are compared.

0.002

Radius (m)

1st Ring Numerical

2nd Ring Numerical
3rd Ring Numerical
1st Ring Experimental
2nd Ring Experimental
3rd Ring Experimental

150

Temperature (oC)

3.2X 10 sin(

900

Temperature (oC)

Heat generation in body is heat transfer to coolant due

to convection
=
Q R 2f l 2 R f lh(Ts Tinf )
Q R 2f
(2.5)
R2
T ( R=
)
(1 2 ) + Ts
Rf
4K f

90

350

280

(2.3)

Apply boundary condition

dT
=0
dr r =0

Axial Heat generation rate

Analytical
Analytical
Numerical
Numerical

380

260

1-D Fourier conduction equation

Parameter
Radius of fuel pin
Length of fuel pin
Fuel thermal conductivity
Fuel density
Fuel specific heat
Inlet Coolant temperature
Heat transfer coefficient at surface

Surface temperature

390

270

1 d dT
r
r dr dr

400

(2.2)

)) + Tin

Q V
x
sin( ) + max 1 cos
+ Tin
l
l
ho A
mc p

QmaxV

boundary conditions for the 3-D code. A good

agreement is found between the numerical code result
with analytical solution obtained using equation 2.2
and 2.3

Temperature (oC)

www.ijnese.org

100

50

0
50

100

150

200

Time (s)

FIG. 9 BENCHMARKING RADIAL FUEL SURFACE

TEMPERATURE PREDICTION OF DIFFERENT RINGS

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

The exercise shows that the multi pin model is able to

capture the rewetting pattern well. However, in this
study, circumferential rewetting pattern could not be
produced as measurements across the circumference
of a simulated fuel pin are not done/reported.
Study on Influence of Rewetting on Conduction
Heat Transfer Assessment
Bottom Re-flooding
The schematic for bottom re-flooding is shown in Fig.
10. The schematic shows the rewetting front travels
from bottom to top.

starting from bottom (0.0 m) to top (3.5 m).

800
700

0.076 m
0.648 m
1.292 m
2.052 m
2.656 m
3.5 m

600

Temperature (oC)

Further, this single pin model has been extended to

multiple pins model considering that pins are
arranged in three concentric rings with radial jet
simulation. A validation exercise has been carried out
with experimental results reported by Patil N. D. The
experimental case with initial temperature of 168C for
all pins with an injection flow rate of 73 lpm has been
considered. The validation exercise results are shown
in Figure 9 at an elevation at 2.4 m.

www.ijnese.org

500
400
300
200
100
0
10

20

30

40

Time (s)

FIG. 11 FUEL TEMPERATURE TRANSIENT AT DIFFERENT

AXIAL LOCATION

The result shows for the chosen boundary conditions

axial conduction is negligible. Cooling by rewetting
does not influence the upstream node temperatures.
The change in upstream node temperature only
happens with the arrival of the rewetting front.
800
r_theta_z
r_theta_z
r_theta_z
r_z_theta
r_z_theta
r_z_theta
theta_r_z
theta_r_z
theta_r_z
theta_z_r
theta_z_r
theta_z_r
z_r_theta
z_r_theta
z_r_theta
z_theta_r
z_theta_r
z_theta_r

700
3.5 m

Temperature (oC)

600
500
2.052 m

400
300

0.076 m

200
100
0
5

10

15

20

25

30

35

40

45

FIG. 12 TRANSIENT SURFACE TEMPERATURE AT DIFFERENT

AXIAL LOCATIONS - A SENSITIVITY STUDIES
FIG. 10 SCHEMATIC FOR BOTTOM REFLOODING

For this study, boundary conditions like re-flooding

velocity is assumed to be 0.35 m/s (Jagdish,2010),
hence over a period of 10 s the rewetting front will
reach the top of the heated fuel pin. As the initial fuel
temperature (750C, Jagdish, 2013) is considered to be
high (as in a practical situation) as compared to Leiden
frost temperature (220C), a linear variation of heat
transfer coefficient from 500 W/m2 K (only steam) to
30000 W/m2 K (only water) and variation of fluid
temperature from 300oC (only steam) to 30oC (only
water) are assumed over 20 s period at the node where
rewetting front has reached. Fig. 11 shows the
transient cooling temperature for six axial locations

As the axial conduction phenomenon is predicted by

numerically technique, influence of ADI solution
technique on the above mentioned findings is
investigated. The ADI solutions are obtained with all
possible directions combinations like r- - z, r- z- , r- z, - z- r, z- r- and z - - r. Fig. 12 shows that the
axial conduction behavior is insensitive to direction of
solution direction. Hence there is no influence of
numerical scheme on the axial conduction behavior.
Radial Re-flooding
The schematic for radial re-flooding is shown in Fig. 13.
The schematic shows the radial direction water jet
travel from centre to the periphery, attempting to
rewet the front and rear surface of fuel pin over the

91

www.ijnese.org

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

minute of the beginning of rewetting of font section.

The influence of one sided rewetting has caused the
maximum temperature of fuel pin to shift from centre
to un-rewetted portion of fuel pin, as shown in Fig. 15.

issuing jet plane.

FIG. 13 SCHEMATIC FOR RADIAL REFLOODING

800
700

r-thita-z
0.004

Temp

500

400
300
200
100

500
450
400
350
300
250
200
150
100
50

0.002

Radius (m)

Temperature (oC)

600

-0.002

0
0

20

40

60

80

-0.004

Time (s)

92

-0.002

0.002

0.004

Radius (m)

FIG. 15 INFLUENCE OF CIRCUMFERENTIAL REWETTING ON

FUEL CENTRE LINE TEMPERATURE

As the circumferential conduction phenomenon is

predicted by numerically technique, and influence of
ADI solution technique on the above mentioned
findings is investigated.
800
700
600
r_theta_Z
r_theta_Z
r_theta_Z
r_theta_Z
theta_r_z
theta_r_z
theta_r_z
theta_r_z
z_r_theta
z_r_theta
z_r_theta
z_r_theta

400
300
200
100

47.25
47.00
46.75

500

TemperatureC)
(

For this study, boundary conditions like high

rewetting heat transfer coefficient of 30000 W/m2 K
(only water) is applied over half of the circumference
(180) of fuel pin where the water jet is likely to
impinge. At the same time, a low heat transfer
coefficient of 500 W/m2 K (only steam) is applied on
the other half of the circumference (180) where the jet
is unlikely to reach. A linear variation of heat transfer
coefficient from 500 W/m2 K (only steam) to 30000
W/m2 K (only water) for the water impinged surface is
considered as described earlier. Variation of fluid
temperature from 300oC (only steam) to 30oC (only
water) is assumed over 20 s period at the node where
rewetting front has reached. Fig. 14 shows the
transient cooling temperature for six circumferential
locations covering the half of circumference which
experiences the water jet. The result shows that for
the chosen boundary conditions, influence of
circumferential conduction is significant. Temperature
of the un-wetted portion is dropped by 250C with
rewetting of the other half of the fuel pin over a

-0.004

Temperature (oC)

FIG. 14 TRANSIENT SURFACE TEMPERATURE ALONG THE

CIRCUMFERENCE OF FUEL PIN

46.50
46.25
46.00
45.75
45.50
31.79

31.80

31.81

31.82

Time (s)

0
0

20

40

60

80

Time (s)
FIG. 16 TRANSIENT SURFACE TEMPERATURE AT DIFFERENT
CIRCUMFERENTIAL LOCATIONS- A SENSITIVITY STUDIES

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

The ADI solutions are obtained with all possible

directions combinations like r- - z, r- z- , - r- z, - zr, z- r- and z - - r as done for axial conduction case.
Fig. 16 shows that the axial conduction behavior is
insensitive to direction of solution direction. Hence
there is no influence of numerical scheme on the axial
conduction behavior.

www.ijnese.org

Q = Local heat generation in fuel (W/m3)

r = Fuel radius changing in r-direction

R f = outer radius of fuel pin (m)
T = temperature (K)
Ts = Fuel surface temperature (K)
Tout =coolant outlet temperature (K)
Tin= coolant inlet temperature (K)

Conclusion

Tinf = Average coolant temperature

Influence of Rewetting on Conduction Heat Transfer

for AHWR Fuel bundle Re-flooding Phenomena has
been studied with the help of 3-D conduction
numerical model. A circumferential conduction is
found to be significant for radial jet rewetting cases as
against a very poor axial conduction during axial
rewetting situation. The conservative assessment
shows that radial jet is able to bring down the unrewetted portion temperature by 250C where the
temperature at different axial location falls only with
arrival of rewetting front. A radial direction jet is
found to be more effective to cool the hot surface as
against the bottom axial reflooding. A sensitivity study
carried out with respect to ADI technique shows that
the solution is insensitive to the direction of solution;
thus eliminating the uncertainty associated with
direction of solution method of ADI.

Tsat = Temperature difference between fuel surface

and saturation temperature of coolant (oC)
V = Volume (m3)

g = Density of steam (kg/m3)

= density fuel (Kg/m3)
= surface tension (N/m)

= angle (radians)
= thermal diffusivity (m2/s)
Lg = density difference between liquid and steam
(kg/m3)
r = distance increment in fuel (m)
z = distance increment in z direction(m)

= angle increment in circumference direction

(radians) t= time (s)
t =time increment(s)

Nomenclature
A = area (m )
C p = specific heat capacity (J/kgK)
2

k f t

Fo =
C p ( r )2

k f t

F =
C p ( )2

k f t

Fz =
C p ( z )2

g = local acceleration due to gravity (m/s2)

gc = gravitation constant (m/s2)

ho = heat transfer coefficient at outer surface (W/m2K)
h fg = Latent heat of vaporization (W/m)

Subscripts
f = fuel
i =mesh point in x-direction
j = mesh point in -direction
k = mesh point in z-direction
Superscript
n= time step
ACKNOWLEDGMENT:

Authors acknowledged Dr. R. K. Singh, Mr. H.G. Lele

and Dr. P. K. Vijayan for their valuable suggestion
during the development of code.
REFERENCES

Blair J.M., An analytical solution to a two-dimensional model

l = length of fuel (m)

of the rewetting of a hot dry rod, Nucl. Eng. Des. 32 (1975)

k g = Thermal conductivity of steam (W/mK)

159170.

k f = thermal conductivity of fuel (W/mK)

Brian, P.L.T. (1961): A Finite Difference Method of Higher-

m = coolant flow rate (kg/s)

Order Accuracy for the Solution of Three- Dimensional

P = Pressure (MPa)

Transient Heat Conduction Problems, AIChE J., 7,

pp.367-370.

93

www.ijnese.org

International Journal of Nuclear Energy Science and Engineering Volume 3 Issue 4, December 2013

Castiglia F., Oliveri E., Taibi S., Vella G., Procedure for

Peaceman, D.W. and Rachford, H.H. (1955): The Numerical

correlating experimental and theoretical results in the

Solution of Parabolic and Elliptic Differential Equations,

rewetting of hot surfaces, Heat Technol. 5 (1987) 8299.

J. Soc. Ind. Appl. Math., 3, pp.28-41.

Chang, M.J., Chow, L.C. and Chang, W.S. (1991): Improved

Raj V.V., Date A.W., Analysis of conduction controlled

Solving

rewetting of hot surfaces based on two region model, in:

Transient Three-Dimensional Heat Diffusion Problems,

Proceedings of 8th International Heat transfer Conference,

Number. Heat Transfer-B, 19, pp.69-84.

San Francisco, California, USA, 1987 1992.

Alternating-Direction

Implicit

Method

for

Coney M.W.E., Calculations on the rewetting of hot

Surfaces, Nucl. Eng. Des. 31 (1974) 246259.
Douglas, J. (1962): Alternating Direction Methods for Three
Space Variables, Num. Math., 4, pp.41-63.

Satapathy A.K., Sahoo R.K., Analysis of rewetting of an

infinite tube by Numerical Fourier Inversion, Int.
Commun. Heat Mass Transfer 5 (2002) 279288.
Sinha R. K, Kakodkar A., Design and development of the

Duffey R.B. Duffey, Porthouse D.T.C., The physics of

AHWR the Indian thorium fuelled innovative nuclear

rewetting in water reactor emergency core cooling, Nucl.

reactor, Nuclear Engineering and Design, volume 236,

Eng. Des. 25 (1973) 379394.

683-700, India (2006).

Duffey R.B., Porthouse D.T.C., Experiments on the cooling

Tien C.L., Yao L.S., Analysis of conduction-controlled

of high temperature surfaces by water jets and drops,

rewetting of a vertical surface, ASME J. Heat Transfer 97

Report No. RD/B/ N2386, Berkeley Nuclear laboratories,

(1975) 161165.

August 1972.
Eckert E.R.G., Drake R.M., Heat and Mass Transfer, second
ed., McGraw Hill, New York, 1959, p. 44.

Thompson T.S., An analysis of the wet-side heat transfer

coefficient during rewetting of a hot dry patch, Nucl. Eng.
Des. 22 (1972) 212 224.

Glasstone S. & Sesonske A. Nuclear Reactor Engineering,

Tyagi Jagdish, kumar Mithilesh, Lele H.G., Munshi P.

Reactor system Engineering volume 2, page no. 565

Thermal hydraulic analysis of the AHWR- The Indian

572.

thorium fuelled innovative nuclear reactor Nuclear

Goodman T.R., The heat balance integral and its application

to problems involving a change of phase, ASME J. Heat
Transfer 80 (1958) 335342.

Engineer and design, volume 262, September 2013, page

21-28.
Tyagi J.P., Kumar Mithilesh, Lele H.G., Munshi, P., 2010.

Jules Thibault A three dimensional Numerical method

Large break LOCA analysis of a natural circulation

based on the superposition principle Numerical Heat

reactor. In: Transactions of ANS Summer Meeting, San

transfer, vol, pp. 127-145, 1984.

Diego, pp. 631632.

Patil N.D., Das P.K., Bhattacharyya S., Sahu S.K. An

Yamanouchi Effect of core spray cooling in transient state

experimental assessment of cooling of a 54-rod bundle by

after loss of coolant accident, J. Nucl. Sci. Tech. 5 (1968)

in-bundle injection Nuclear Engineering and Design 250

547558.

(2012) 500 511.

94