Neutron Transportation theory

One-speed (mono energetic)

diffusion equation
we got: 
       1  ( r , t)
divD( r ) ( r , t) - a ( r )  ( r , t) + f f ( r)  (r, t) + S( r , t) =
V t
This is true where the Fick’s Law approximations are true
1) the stationary task (there is no dependence on temporary variable):
      
divD(r ) ( r ) -  a ( r )  ( r ) + f  f ( r )  ( r ) + S( r ) = 0
2) the stationary task, the medium isn't multiplying (section of fission is equal to zero):
    
divD( r ) ( r ) -  a ( r )  ( r ) + S( r ) = 0
3) homogeneous (properties of the environment don't depend on space variable) not
multiplying medium, the stationary task
 1  1 
( r ) - 2   ( r ) + S( r ) = 0
L D
As in case of the homogeneous medium the first item the diffusion coefficient can be carried out for a
divergence sign, and the divergence from a gradient of function is a Laplacian Δ of this function.

2
 Initial and boundary conditions (2)
In case of one dimension:

3
 Initial and boundary conditions (1)

I. Function of neutron flux isn't negative and limited.

II. As usual in an initial time point of t=0 function of neutron flux is
known.
III. Conditions on interface of two mediums with different properties
+  + 
i ( rs , t) = i n,2 ( rs , t)
n,1
Medium 1 Medium 2 -  - 
D1, 1 D2 , 2 i n,1 ( rs , t) = i n,2 ( rs , t)
 After simple transformation:
dS = 1 n
 
 1 ( rs , t) =  2 ( rs , t)
rs
 
D1 n 1 ( rs , t) = D2 n  2 ( rs , t)
4
 Initial and boundary conditions (3)
IV. Conditions on the boundary medium - vacuum

 1  D 
Medium
D, 
Vacuum
i ( rs , t) = 0 or  ( rs , t) +  n  ( rs , t) = 0
-
n
4 2
 
dS = 1 n  n ( rs , t) 1 3 1
so:  = - = -  = -

rs
( rs , t) 2D 2
tr


where the complex standing in the left often is called a logarithmic derivative,
and constant

2
  tr - length of linear extrapolation of neutron flux to vacuum.
3

5
 Initial and boundary conditions (4)
Slab geometry: The flux in the medium is
distributed as it is shown in drawing.

Vacuum
How the neutron flux in vacuum will behave?

6
 Initial and boundary conditions (5)
1 d 1
 .
1 d1 d( x s ) 1dx x  xs 1 
1  d.  . 1
   .
Real flux( x s ) dxs x 0xs x xs
( x ) d
x -? x
( x s ) dx x  xs 

Vacuum
x0 -?x0 -? C1 x0  C2  0
x0 -?
C x  0C2  0
C1 x0 1 C02 Получим: C1 1
C1 x0 CCC2  0 
C x  C2 1  '
1 11 s
Получим: 1
Получим:  C  1
 1=  
x  C12 s  ' 2 0 s '  
Таким C
CПолучим: x
образом,C x x +
Linear 1 s
C x случае:
C2
 '
Таким образом, x0 = xs + 0 s
Таким или в
образом,более x =
общемx +
1 s
extrapolation
to zero или в Такимобщем
более образом,случае:x0 = xs + 
или в более общем  случае:
 (   n, t)общем
илиrsвболее  0 случае:
2 
  tr  0,66  tr (r Ingeneral  case:
3 
( rs  s n, t)  0, t)  0
n
xs x0 ( rs    n, t)  0
7
 Initial and boundary conditions (6)
Precise solution

Improved
diffusion

Elementary
diffusion
  0,71 tr

8
 Initial and boundary conditions (7)
Уравнение диффузии для случая стационарной задачи
  V. Condition  the localized  source   
) + f Уравнение
- div i ( r ) -  a ( r )  ( rof ( r )   ( r ) +
диффузии S 
для ( 0 ) = 0,
r - rстационарной
случая задачи
f       
- div i ( r ) -  a ( r )  ( r ) + f f ( r)  (r) + S   ( r - r0 ) = 0,
Surround a source
Уравнение диффузии для withслучая стационарной задачи
R  with the  sphere  R radius  the center in
  a
   integrate
  source, - div i (on r )  a volume
-the ( r ) ( rof ) +thef sphere
( r )  (r ) + S   ( ) = 0,
r - r0 at
 
and take a limit
- lim R 0  d r div i ( r )  - lim R 0  dr   a ( r ) f
 (r   
) + 
R 0    R 0   r  a ( r ) ( r )  +
convergence
- lim of radius d r of
diva i sphere
( r ) - to
lim zero: d
V  V   V
   
  V
   
    - lim R0  d r divi ( r )- lim R 0 d r  a ( r ) ( r )  +

V

+ lim R 0  d r  f  f ( r ) ( r+)lim  R+0  
lim 
d
 V   V    
rV 
Rf  f0( r )  d( 

r r
)  S
+ lim R 0 
(Vr
V
d -
r Sr   )(
= 0
0 r - r0 ) = 0

  
+ lim R 0  d r  f  f ( r ) ( r ) + lim R 0  d r S   ( r - r0 ) = 0
r0
V     V
   lim R 0  drs i n ( rs )  S
 Получаем:
Получаем:
Г

Hence: lim R 0  d rs i n ( rs )  S  Г 
 lim R 0  drs i n (rs )  S
 
Получаем:
In the left part of the formula there is an expression which  Гis equal to number of neutrons
which cross in unit of time a surface of the sphere of infinitesimal radius surrounding a
source.
It is natural that this number of neutrons is equal to the power of an external source.
9
 Problem statement
Stationary problem for the homogeneous non-multiplication medium. Find Ф(r)

 1  1 
( r ) - 2   ( r ) + S( r ) = 0.
L D
Solution:
1. Write the general equation in relation to specific geometry of the medium (a
Laplacian in the first member and variables on which the solution depends) and to a
specific type of an external source;
2. Finding of the common solution of a differential equation of the second order;
3. Determination of the unknown constants :
a) The flux is non-negative and limited, possesses property of symmetry;
b) Condition on interface of two mediums;
c) Condition on outline border the medium-vacuum;
d) Condition of the localized source.

10
 Laplace’s operator

11
 Symmetries

12
 Solutions of Laplace equation
Symmetry Slab Cylinder Sphere

1 d  d (  )  1 d  2 d (r ) 
d 2 ( x)  d  

d 
 r 
Laplace equation   2
( x)  0 r dr 
2
dr 
dx 2   2  (  )  0   2  (r )  0

exp(  r )
 (r )  C1 
 2   2 ( x)  C1  exp(  x) (  )  C1  I 0 (   ) r
C2  exp(  x) exp(  r )
C2  K0 (   )  C2 
r
Solutions
sin(  r )
( x)  C1  sin(  x) (  )  C1  J 0 (   )  (r )  C1 
r
2 2 C2  cos(  x) C2  N0 (   ) cos(  r )
 C2 
r
13
 Task #1 – Point source in infinite
spherical geometry (1)
Point source q power [1/s] in the infinite, non multiplying, homogeneous
medium. To find the Φ(r) function.

1 d 2 d 1 1
2
r  (r) - 2   (r) + q (r) = 0, r  0 .
r dr dr L D
for r>0

1 d 2 d 1
2
r  (r) - 2   (r) = 0, r >0.
r dr dr L
Common solution:

exp - Lr  exp Lr 

(r) = C1 + C2 .
r r
C2  0. Neutron flux is limited
14
 Task #1 – Point source in infinite
spherical geometry(2) .

Localized source condition:

    i
: lim R 0  d rs i n ( rs )  q
Г 

  d(r) d(r)  exp- RL  exp- RL  R

i n (rs )  (n  i ( rs )) = -D   -D r =R  DC1  2
  .
dn dr  R LR 

   
 
lim R 0  d rs i n ( rs ) = lim R 0 4R 2  i n (R) =
S 
  
lim R0 4DC1 exp- RL  exp- RL    4DC1  q
R
  L 

q q exp- Lr 
C1   (r) = .
4D 4D r
15
 Task #1 – Point source in infinite
spherical geometry(3)
Ф(r) .

2  3tr
q exp- Lr 
 (r) = .
4D r

r
16 0
 Physical meaning of diffusion length
square
Find r 2- an
Найдем величину average квадрат
- средний square ofсмещения
shift of a neutron (on(по прямой) от точки
нейтрона
рождения доa точки
straight line) from a birth point to an absorption
поглощения.
point


2
dV r p(r)
r2 = V where : p(r) = A  a (r)
 dV p(r)
V
 
4  drr r A a ( q
2 2
)r exp-
-1 r   dr r 3
exp - r

4D L L

r2  0

= 0

= 6 L2
4  drr 2 A a ( q )r -1 exp- r   dr r exp-  r
4D L L
0 0

L2 makes the following physical sense: the square of length of diffusion of

neutrons is one sixth average squares of shift (on a straight line) a neutron from
a birth point to an absorption point.
17
 Task #2 – Plane source in infinite
planar geometry (1)
Planar source the power of q [1 / (sm2 s)] in the infinite, non
multiplying, homogeneous medium

d2 1 1
2
 (x) - 2
  (x) + q (x) = 0
dx L 2D

x >0
d2 1
2
 (x) - 2
  (x) =0
dx L

: ( x) = C1 exp- Lx  + C2 expLx .

C2  0. Flux is limited
18
 Task #2 – Plane source in infinite
planar geometry (2)
Localized source condition:

q q
lim  0 i x ( )  ,
2
d (x)
i x ( )  -D .
dx
x =

d (x)
 - 1 exp- Lx 
C
dx L

exp- L  
DC1 DC1 q qL
lim  0  . C1 
L L 2 2D

exp- Lx 
qL
 ( x) =
2D
19
 Task #3 – Thread-like source in
infinite cylindrical geometry (1)
Thread-like source q power (1/cm s) in the infinite, non multiplying,
homogeneous medium.

1 d d 1 1
 (  ) - 2  (  ) + q (  ) = 0,
 d d L D
1 d d 1
асти источника (>0):  (  ) - 2   (  ) = 0, >0.
 d d L
1 d d 1
  (  ) - 2   (  ) = 0,
 d d L

 (  )  C1 K 0 ( L ) + C 2 I 0 ( L ) , где K 0 ( x)

C2  0. Flux is limited

20
 Task #3 – Thread-like source in
infinite cylindrical geometry (2)
   
Localized source condition: lim R 0  d rs i n ( rs ) = q .
S 

  d(  ) d(  )  dK 0 ( L )  DC1

i n (rs )  (n  i ( rs )) = -D   =R  -D  =R  DC1    R = K 1 ( RL ).
dn d  d   L

dK 0 ( x )    
dx
= - K1 ( x ) ,
lim R 0  d rs i n ( rs ) = lim R 0 2R  i n (R) =
S 
 C  R 
= lim R 0 2RD 1 K1 ( RL )  2DC1 lim R 0  K1 ( RL )   2DC1 = q.
 L  L 

lim x0 ( xK1 ( x))  1. q q

C1  (  )  
K0 ( L ).
2D 2D
21
 Task #4 – Plane source in finite
planar geometry (bare slab) (1)
Real
Problem: find neutron
boundary
flux from plane source
with S power
(neut./cm2 s) in a plate
S (n/сm2 s) 
thickness 2a
vacuum medium medium vacuum

Extrapolated
boundary
aэ  a  
22
 Task #4 – Plane source in finite
planar geometry (bare slab) (2)
Acquisition of solution:
1
(1) Compile the diffusion equation in the medium (x>0) ( x)  2 ( x)  0
L
x x
 
(2) Write the common solution ( x)  C1e L
 C2e L

(aэ )  0
(3) Write the boundary conditions S
in ( ) 
 0

2
(4) Get the solution
 aэ  x 
sinh  
S L  L 
 ( x) 
2 D  aэ  ,where
cosh  
L
23
 Task #5 – Distributed source in bare
slab (1)
Extrapolated
Problem: find neutron flux from boundary
distributed source with S power
(neut./cm3 s) in a plate thickness 2a
medium medium aэ  a  

S (n/сm3 s)
vacuum vacuum

x
0

1 S
( x)  2
 ( x )  0
L D
24
 Task #5 – Distributed source in bare
slab (2)
Acquisition of solution: 1 S
(1) Compile the diffusion equation in the ( x)  2
 ( x )  0
medium (x>0) L D
x x
 S 
(2) Write the common solution ( x)  C1e  C2e 
L L
a
Common Particular
Uniform sol. solution

(3) Write the boundary conditions

(aэ )  0.
( x  0)  0.

 x 
 cosh   
(4) Get the solution S
 ( x)  1  L 
a   aэ  
 cosh  L  
25
  
 Task #5 – Distributed source in bare
slab (3)
 
S  1 
1  
Solution: a   aэ  
 cosh   
  L
 x 
 cosh   
 ( x) 
S
 1  L 
a   aэ  
 cosh  L  
  

Question: How will change the

distribution of neutron flux if we
increase the size of a plate (a), that
is a >> L ?
-aэ aэ
26
 Task #5 – Distributed source in bare
slab (4)
Solution S
a
 x 
 cosh   
 ( x) 
S
 1  L 
,
a   aэ    
 cosh  L    
   S
1 
a 
1

 aэ  
 cosh  L  
S   
at aэ   ( x) 
a

For material fixing: try to draw

approximately as functions behave

in ( x); in ( x); in ( x).

-aэ aэ
27
 Principle of superposition of sources
If in the medium there are not one source of neutrons but several neutron sources, that
resultant neutron field will represent superposition of neutron fields from each of sources
as at the description of the neutron field interaction of neutrons among themselves isn't
considered

Example: if in infinite homogeneous

 environment there are the I point sources of
r
 *q1 neutrons of various power, then the neutron
r1 *q2 flux in a point r will be defined by expression:
 
r2 r3 *q3
.
.
.


 
rI *qI
r-ri
1 I exp

L
( r) = qi
4 D i=1 r-ri

28
 Solution in a finite medium
If the medium is finite, there is the leakage additional to absorption. To compensate it the
following statement have to be true
 f  f  a 
 f  f  a or    B 2
0
 D 
Thus for the equation:

 f  f   a 
( x)     ( x)  0; ( x)  B2 ( x)  0
 D 
Common solution of the equation will be combination of sinus and cosine:

( x)  C1  cos( B  x)  C2  sin( B  x)

29
 Solution in a finite medium,
geometrical and material parameters
From symmetry condition С2 = 0.

From boundary condition medium-vacuum:

(aэ )  C1  cos( B  aэ )  0.
 
B  aэ  Hence: B
2 2  aэ
 f  f   a    
2

  
 D   2  aэ 
Material Buckling
parameter

This, so-called, a formula of equality of material

and geometrical parameter (buckling) of the
medium - if It is true, then the reactor is critical -a aэ
э

30
 Example: Solution in 3-D geometry
z Cubic reactor with size equals a=80 cm. Reactor
homogeneous with following parameters:

 f  f  0.242 cm1 ;
D  1 cm.
Find macroscopic cross section of absorption
a to criticality
a
0

y
x

31
 Acquire solution (1)
Write diffusion equation. The neutron flux depends on three variables: x,y,z

 f  f  a 
 x , y , z ( x, y, z )     ( x, y, z )  0
 D 
From Math It is well known that we can split variables as following:
 ( x, y, z) = X( x)Y ( y)Z(z) .

X ( x) Y ( y) Z ( z )  f  f  a 

      3  B 2

X ( x) Y ( y ) Z ( z )  D 

Thus: X ( x)  C1  sin( B  x);

Y ( y )  C2  sin( B  y );
Z ( z )  C3  sin( B  z ).
32
 Acquire solution (2)
z
Criticality condition (material parameter
and buckling are equal) for cube :

 f  f   a    
2

   3 
 D   aэ 
a
Hence:  a  0.237 сm 1
0
 f  f  0.242 сm1 ;
y
D  1 сm.
x
Critical volume of cube is:

Vcube  a3  0.512 m3
33
 Question
At the same parameters of the medium compare critical volumes of the reactor in the form
of the sphere of radius of R and the cylinder of radius R and height of H.

 a  0.237 сm 1
 f  f  0.242 сm1 ;
D  1 сm.

Cube Sphere Cylinder

0.512 m 3 ? ?

34
 Summary for different geometries

35
 Microscopic cross section of
scattering
As usual there are three regions in dependency microscopic cross section of scattering to
energy: potential scattering, resonances (exclude hydrogen and deuterium) , smooth region

37
 Microscopic cross section of
scattering (2)
U-238 He-4

H-1 Be-9

38
 Considering Slowing down energy
region

El. Scat. cross section doesn’t

Eleft ~ 1 eV depend on energy Eright ~ 100 keV

Slowing down region or Resonance

region or
Thermal Region of Potential elastic Potential elastic
region scattering (isotropic in scattering which
is anisotropic in
CoM) CoM
10 MeV
E max 
A3 2
39
 Additional assumptions for slowing
down energy region (1 eV to 100 keV)
I. Neutron impacts to single nucleus (neutron doesn’t
feel the nuclei binding in molecule or crystal lattice)
• E= 0.025 eV v= 2.2×103m/s λ = 1.8 Å
• E = 10 keV v=1.4×106m/s λ = 0.0455 Å
• E = 1 MeV v= 1.4×107m/s λ = 0.000287 Å
Size of atom ~ 1 Å
Size of nucleus ~ 10-4Å
Distance between atoms in solids ~ few Å
II. Before scattering the nucleus is static in lab coordinates
(the neutron speed much more to thermal oscillatory
movement of nucleus
m/s
Neutron eV

m/s
40
 Elastic scattering Law
Problem: find distribution of neutron energy
in post collision energy range

E1 – the energy before impact

E2 – the energy after impact dE

P(E1→E2) (Elastic scattering law) - the E2

density of probability that neutron after
E1  E2  E1 E2  E1; E1 
collision will have the energy E2 within post
collision energy range или
Post collision energy range
P(E1→E2)*dE - probability that neutron after
collision will have the energy within range dE
near E2

41
 Acquire solution (1)
p  E1  E2  dE2  W   d d
a
d
p  E1  E2   W   b 
dE2 r
O
Sring 2 R 2 sin d 1
W   d    sin d
4 R 2 R
S sphere 2
b1
Sring  2  r  ab  2  R sin  Rd  2 R sin d
2 a1

Neutron scattering in CoM is

Ssphere  4 R 2 isotropic

Verification:
W    sin 
1
 
2
0 dW 
  
1
 d sin  
1
2 1
20 2

42
 Acquire solution (2)
A  2  A cos  1  A  1
2
E2 A2  1
2
E2  E1 cos  
 A  1
2
2A E1 2A

 A  1 1  A  1
2 2
d
 sin d  dE2 
2 AE1 dE2 sin 2 AE1

d 1  A  1  A  1 1
2 2
p  E1  E2   W   1
p  E1  E2   sin 
dE2 2 sin 2 AE1 4 A E1

 A 1 A2  2 A  1  A2  2 A  1
2
4A
1  1    
 A  1   A  12
 A  12

1
p  E1  E2  
1    E1
43
 Acquire solution (3)
1 p(E1  E')
p  E1  E2  
1    E1

Verification: 1
E1 (1 )
E1
1
Y   dEp  E1  E   E1 1     1.
 E1 E1 1   

E
E1 E1

Thus, the elastic scattering law means that after collision neutron at energy E1 to nucleus
with mass A the it’s post collision energy E2 equally probable distributes within post
collision energy range

44
 Average post collision energy

Average post collision energy p(E1  E')

1 
E2  E1
2
1
Average post collision energy loss E1 (1 )
1
E  E1
2 E
E
E1 E1

E2
45
 Average cosine in CoM
 
1
 d  cos  W   0 d  cos   
2
sin  
cos    0

 
0
1
 d W  
0
0 2 sin  
d 



2

n 

2
Impact

46
 Average cosine in Lab coordinates

1  A cos
f    cos    d 
1
 sin  
0 1  A  2 A cos 2
2

1 A x
1
1
  dx  , где x  cos .
2 1 1  A2  2 A  x

2
cos() 
3 A

47
 Average cosine in Lab coordinates

  cos()
2
cos() 
3 A

48
 Elastic Scattering from Molecules

Free and bound state on

Hydrogen nucleus

49
 Average logarithmic energy loss

50
 Balance equation
Take infinite, non multiplying, homogeneous medium in which there are isotropic,
stationary external uniformly distributed sources emitted neutron with energy E0
Lets find the neutron balance in small range dE near energy value E

 PS E   AE   RS E   QE   0
where
PS E  – disappearance neutron rate by elastic scattering;
AE  – disappearance neutron rate by absorption;
RS E  – appearance neutron rate in dE by elastic
scattering from upper energies;
QE  – appearance neutron rate by outer sources.

PS E    S E E dE

AE   a E E dE

QE   q E  E0 dE
52
 Appearance neutron rate by elastic
scattering from upper energies
dE dE'
E0 E0
  E
E
E E' 
Neutron ca appear in dE only by scattering from [E, E/α] (upper post collision range). Take
any small range dE’ from [E, E/α]

 S EEdE - the amount of neutron scattered in dE’ per unit of time

dE
- the possibility for neutron to appear in dE after scattering in dE’
1   E 
dE - the amount of neutron scattered in dE’ and after that
S  E    E  dE 
   having the energy in dE per unit of time
1    E
To find Rs(E) we have to integrate from E to E/α or E0 (what less)
53
 Slowing down equation
 E
min  E0 , 
 
 S  E     s ( E ')
 S  E    E    a  E    E    dE   q  E  E0   0
E 1    E
Additional quantities:

FS E   S E E  - scattering density

F E   tot E E  - collision density

 E
min  E0 , 
1  
FS  E 
F E   dE    dE q  E  E0 
1 E
E

54
 Slowing down density
j(E)
dE'
E0 E0
  E
E' E E' E

Slowing down density - describes amount of neutrons in the single spatial volume which
energy during slowing down in unit of time changes from value above, to value below the
energy E or cross the energy E per unit of time during slowing down. This function is called –
Slowing down density and is designated as j(E).

 E
min  E0 , 


E  E 
j E   E dE FS E 1   E
neut.
 j  E   
sec
55
 Slowing down density (2)
E0

j E    dE  FS E 
E
In case of hydrogen (α=0):
E
E

In case of infinite, homogeneous, non-absorbing medium the slowing down density is

constant and equals to power of outer monochromatic source

j E  q

In case of absorbing medium the slowing down density decreases during the process
of slowing down. Derivation of the corresponding slowing down densities at E1 and Е2
(E1 > Е2 ) equals so important in nuclear reactor physics quantity as possibility to
avoid of absorption during slowing down from E1 to Е2 .

j  E2 
  E1  E2  
j  E1 
56
 Slowing down on Hydrogen
without absorption
Take infinite, non multiplying, homogeneous, non-absorbing medium consisted of nuclei of
hydrogen in which there are isotropic, stationary external uniformly distributed sources
with power q [n/s] emitted neutron with energy E0
Slowing down equation for the medium:
dE 
E0

 FS E    FS E   q E  E 0   0
E
E
Skip the solution of integral equation

 E  
q
- Fermi spectrum for
 S E   E energy variable
Hence:

In lethargy variable:   E  dE    u  du  u     E   E
q
 u   - Fermi spectrum for
S  u 
Hence:
lethargy variable
57
 Slowing down on Hydrogen
without absorption

Find slowing down density for this problem:

dE  q
E0 E

0

j E    FS E  dE   E 
E
  q  E   E 
0 
E
E E
E  E 
 1 E0  q  1
E0
 q
 qE  E  qE  E q

 E E  E0  E E  E0 .

Slowing down density for non absorbing medium is constant and

equals sources power q
58
 Slowing down on nuclei with A>1
without absorption
Take infinite, non multiplying, homogeneous, non-absorbing medium consisted from any
nuclei with A>1 in which there are isotropic, stationary external uniformly distributed
sources with power q [n/s] emitted neutron with energy E0
Slowing down equation for whole energy
interval:
 E
min  E0 , 
1  
FS  E  
F E   dE    dE 
1 E
E
 q  E  E0 
The solution too hard as it did for
hydrogen medium. In this case we should
to find solution separately in every
collision interval

59
 Slowing down on nuclei with A>1
without absorption
Slowing down equation for the 1st post collision
energy range (collision interval) :

FS1 E 
E0

 FS E    dE 
1
 q E  E 0   0
E
1   E 
Solution: for E   E0 , E0 

q
E01
F 1  E   1  
S 1
1
E
For the 2st post collision energy range the solution is complicated
But it is simply to show there is discontinuity of the function of
scattering density Fs(E) at the point αE on value:
q
1   E0
60
 Collision density in energy

q
 asympt  E  
 E S  E 

61
 Collision density in energy – Placzek
function

62
 Slowing down in mixtures
Take infinite, non multiplying, homogeneous, non-absorbing medium consisted of mixture of
isotopes (Ai, Ni, σi) in which there are isotropic, stationary external uniformly distributed
sources with power q [n/s] emitted neutron with energy E0
Considering only asymptotic energy range:
E

1 i  Si  E    asympt  E  
  Si  E   asympt  E     dE 0
i i 1  i E
E
Solution: q
 asympt  E  
  E S  E 
q - the Fermi spectrum is formed
 asympt  u  
  S  u 
 Si
where   i
i S

63
 Neutron thermalization
Scattering process of neutron on nuclei in thermal energy range (from 0 to ~ 1-5 eV) is
called Thermalization
Features of thermal range:
1. The kinetic neutron energy is comparable to thermal oscillation movement of
nuclei – 3/2kT .
2. De Broil wave length of neutron is comparable to molecule size or crystal
structures

For most isotopes the cross section in thermal range

behaves as 1/v Law (for thermal neutrons (in 1/v
region), absorption cross sections increases as the
velocity (kinetic energy) of the neutron decreases)

64
 Balance equation in thermal range
Balance equation for phase volume (r,E):
Eth
   
D(E) Ф M (r , E )  tot ( E )ФМ ( r , E)   dE'  S ( E ' )ФМ ( r , E' )p(E'  E)  S(r , E)  0
0
Integrate by energy for all thermal range and take into consideration new quantities: the
averaged cross section and thermal flux
Eth Eth

 dE a ( E )ФМ (E)  dED( E )Ф М (E) Eth

a  0

ФТ
D 0

ФТ
ФТ   dEФ
0
М (E')

Hence:
Eth
D  Ф Т (r)- a  ФТ (r)   dES(r,E)  0
0
Eth  1  1 
Ф( r ) - 2 Ф( r )  j( r , E th )  0
j(r,E th )   dES(r,E)
0
LТ D
D
L 
2
- Diffusion length square of
Slowing down density a thermal neutrons
65
 Balance equation
The energy dependence of the flux was discussed and a method introduced
for approximating the dependence for the thermal flux. Since neutrons in
a nuclear reactor actually have a distribution in energy, this distribution must be
accounted for in the diffusion equation.
The balance equation in phase space (r,E)
 Emax
  

    
 div i r , E tot r , E  r , E   
 S
 r, EW r;EE r, EdE S r, E 0
 E
 
 

  
 grad  r , E  3tr r , E i r , E

  

 1

 1   E  , E   E , 
 E  - scattering indicatrix for the potential elastic
W r , E   E 
 scattering. In general It is much more complex

 0, E  E
 

66
 Multi group method
One of the most effective ways to calculate the slowing down and
diffusion of neutrons is by the group-diffusion method. In this
method, the entire range of neutron energy is divided into N
energy intervals as indicated in the figure →
All of the neutrons within each energy interval are
then lumped together, and their diffusion, scattering, and
absorption are described in terms of suitably averaged
diffusion coefficients and cross-sections.
New quantities:
n r    nr , E dE
  Group g
Integrated neutron density for the group g: g
E
g
 nr , E dE

E
  r

, E dE   n r

, E v  E dE 
g
 v  n r    r 
 
 nr , E dE
 g g g
E E
g g E
g

 (r ) - Neutron flux in group g.
g
67
 Averaged cross section
 
 
tot
r , E  r , E dE
E
    
  r , E  r , E dE 
g
    r , E dE   r  r 
1.
E
tot    r , E dE E
g g
g E g
g

 (r ) -macroscopic cross section of neutron interaction (total) in group g
g
(group total cross section)

 dE   S  r , E  W  E   E    r , E   dE 
E E
g g g  g
2.    r , E   dE    S r  g ' r 
   
r , E  dE 
E
E g'
g'
g '  g  -macroscopic cross section of transfer neutron from g to g’
 (r )
s (group transfer cross section)

68
 Diffusion equation for the group g
N
g  g'
 s  (r )  g (r ) Total transfer neutron rate out of g
g  g 1

g 1
g' g
 s  (r )   g  (r ) Total transfer neutron rate into g
g 1

 gg 
 r    r    r 
g 
a g S macroscopic cross section of absorption in group g
g g g G g  g (group absorption cross section)
     
a c f S
g  g 1
Combining the prior terms in the obvious way gives the following steady state
diffusion equation for the gth group neutrons :

 g 1 g  g 
div D r grad  r    r  r     (r ) r   q r   0
  g   
g g a g g g

g 1
69
 Multi group diffusion approximation
Emax
divD  r , E   grad   r , E  tot  r , E    r , E      r , E  W  r ;E  E   r , E  dE   S  r , E   0
S
E
 g 1 g  g 
div D r grad  r    r  r     (r ) r   q r   0
  g   
g g a g g g
g  1
Yet, this two equations are absolutely the same. The question how to find
group cross sections?

Suppose, that within neutron energy group g (energy range ΔEg ) the neutron field has
uniformly spatial distribution . Thus we can split variables:
nr , E   R g r S g E 
 

     E  E g g  1,2,..., G
i r , E   I g r S g E  

S ( E ) - intragroup spectrum
g
Multi group diffusion approximation – means that a priori known spectra (Maxwell’s,
Fermi etc) are taken as intragroup spectrum
70
 Intragroup spectrum
S(E)
Slowing down – Fermi spectrum

Fission
spectrum

Maxwell’s
spectrum

N g 1 Е

~ 1 eV Е0=15 МeV
71
 Averaged microscopic cross sections

By using formula for  (rand
) Ф(r,E) and n(r,E):
g
  tot E v( E ) S g E dE
E
 r  
g  g
tot  v( E ) S g E dE
E
g

Taken from Intragroup

JANIS spectrum

The same:

 dE  v( E ' ) S E S g E WS r , E   E dE    tr E S g E dE
g   g E g E g  g
E
g
  ,  
s  v( E ' ) S g E dE  tr  S g E dE
E E
g g

72
 Multi group equations
EG Eg E1

 
E
0 G E g 1 Emax
Group
Номер number группы
энергетической

 g 1 g  g 
div D r grad  r    r  r     (r ) r   q r   0
  g   
g g a g g g
g  1
In the end we got the system of differential equations of second orders without dependency of
variables to energy. The amount of equations equals to amount of energy groups.
In a number of reactor calculations, especially those involving the criticality of thermal reactors,
at least two groups of neutrons must be used to obtain reasonably accurate results.
The simplest case with one energy group is one speed diffusion equation (diffusion of mono
energy neutrons)

73
 Multi group equations
E G Eg E1

     
E
0 G E g 1 E max
Номер энергетической группы
Group number

 g 1 g  g 
div D r grad  r    r  r     (r ) r   q r   0
  g   
g g a g g g
g  1

Because of the complexity of the group-diffusion equations, it is common practice to

use a computer program to evaluate the group fluxes. The techniques by which this is done
involve approximating the derivatives by numerical methods and then requiring the equations
hold only at a series of discrete points in space. The equations are then reduced to algebraic
equations valid for only those points. The exact approaches vary, and such methods lie beyond
the scope of this material. It suffices here to note that many computer programs such have been
written by which the equations represented above can be evaluated.

74
 Example: Three groups diffusion
equations

  
D  r   1 Φ r   q r   0
1 1 a 1 1
  1 2  
D  r     r   
2  r   q r   0
2 2 a 2 s 1 2

  1 3  23  
D  r     r   
3  r     r   q r   0
3 3 a 3 s 1 s 2 3

1 1 1 1  2 1 3
     13
a c f s s s
1 2
2  2  2  2  3
s
a c f s 2
s
3

E
3  3  3 g 3 g 2 g 1 E
a c f 0 3f 2f 1f
max
3c c2 1c
75