Chapter 3.

Deriving the Fluid Equations From the Vlasov Equation 25

Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Topics or concepts to learn in Chapter 3:

1. The basic equations for study kinetic plasma physics: The Vlasov-Maxwell equations
2. Definition of fluid variables: number density, mass density, average velocity, thermal
pressure (including scalar pressure and pressure tensor), heat flux, and entropy function
3. Derivation of plasma fluid equations from Vlasov-Maxwell equations:
(a) The ion-electron two-fluid equations
(b) The one-fluid equations and the MHD (magnetohydrodynamic) equations
(c) The continuity equations of the number density, the mass density, and the charge density
(d) The momentum equation
(e) The momentum of the plasma and the E-, B- fields
(f) The momentum flux (the pressure tensor) of the plasma and the E-, B- fields
(g) The energy of the plasma and the E-, B- fields
(h) The energy flux of the plasma and the E-, B- fields
(i) The energy equations and the equations of state
(j) The MHD Ohms law and the generalized Ohms law

Suggested Readings:
(1) Section 7.1 in Nicholson (1983)
(2) Chapter 3 in Krall and Trivelpiece (1973)
(3) Chapter 3 in F. F. Chen (1984)

3.1. The Vlasov-Maxwell System

Vlasov equation of the th species, shown in Eq. (2.7), can be rewritten as

f (x,v,t) e
+ v f (x,v,t) + [E(x,t) + v B(x,t)] v f (x,v,t) = 0 (3.1)
t m
or
f (x,v,t) e
+ {vf (x,v,t)} + v {[E(x,t) + v B(x,t)] f (x,v,t)} = 0 (3.1')
t m
where = / x and v = / v .
26 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Maxwells equations in Table 1.1. can be rewritten as

c (x,t)
E(x,t) =
0
1
= e n (x,t) (3.2)
0

e
1
=
0
f (x,v,t)d 3v

B(x,t) = 0 (3.3)
B(x,t)
E(x,t) = (3.4)
t
E(x,t)
B(x,t) = 0 J(x,t) + 00
t
E(x,t)
= 0 [ e n (x,t)V (x,t)] + 00 (3.5)
t
E(x,t)
= 0 [ e vf (x,v,t)d 3v ] + 00
t

The Vlasov equations of ions and electrons and the Maxwells equations are the governing
equations of the Vlasov-Maxwell system, which includes eight unknowns ( f i, f e , E, B ) and
eight independent equations in a six-dimensional phase space.
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 27

3.2. The Fluid Variables

Before introducing the fluid equations, we need to define fluid variables of plasma.
The number density of the th species, in Eq. (3.2), is defined by

n (x,t) f (x, v,t)d 3v (3.6)

The average velocity of the th species, in Eq. (3.5), is defined by

V (x,t)
vf (x, v,t)d 3v
(3.7)

n (x,t)
The particle flux of the th species is

n V = vf (x, v,t)d 3v (3.8)

The mass flux of the th species is

m n V = m vf (x, v,t)d 3v (3.9)

The charge flux of the th species is

e n V = e vf
(x, v,t)d 3v (3.10)

The momentum flux, or the kinetic pressure, of the th species is

m n V V + P (x,t) = m vvf
(x, v,t)d 3v (3.11)

where m n V V is the dynamic pressure, and P (x,t) is the thermal pressure tensor.
The thermal pressure tensor P (x,t) in Eq. (3.11) is defined by

P (x,t) m [v V (x,t)][v V (x,t)] f

(x, v,t)d 3v (3.12)

Since P (x,t) is a second rank symmetric tensor, trace of P (x,t) is invariant after an
orthonormal coordinate transformation. For an isotropic pressure, we have
p (x,t) 0 0

P (x,t) = 1 p (x,t) = 0 p (x,t) 0

0 0 p (x,t)
Thus, in general we can define a scalar thermal pressure p (x,t)
1
p (x,t) trace[P (x,t)] (3.13)
3
The flux of the total kinetic pressure of the th species is

m n V V V + (P V )S + Q (x,t) = m vvvf
(x, v,t)d 3v (3.14)

where the heat-flux tensor Q (x,t) is a third rank tensor, which is defined by
28 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Q (x,t) m [v V (x,t)][v V (x,t)][v V (x,t)] f

(x, v,t)d 3v (3.15)

(P V )S in Eq. (3.14) is a symmetric third rank tensor, which is defined by

(P V )S P V + V P + m (v V )V (v V ) f (x, v,t)d 3v (3.16)

The kinetic energy flux of the th species is

1 3 1
2
m n V V V + p V + P V + q (x,t) =
2
2 m v vvf (x,v,t)d 3v

or
1 3 1
( m nV2 + p )V + P V + q (x,t) =
2 2
2 m v vf
2
(x,v,t)d 3v (3.17)

where the heat-flux vector q (x,t) is defined by

1
q (x,t) 2 m [v V (x,t)] [v V (x,t)][v V (x,t)] f
(x,v,t)d 3v (3.18)

For advanced study:

If it is needed, we can define a fourth rank tensor R (x,t)

R (x,t) m (v V )(v V )(v V )(v V ) f (x, v,t)d 3v (3.19)

and a fifth rank tensor S (x,t)

S (x,t) m (v V )(v V )(v V )(v V )(v V ) f (x, v,t)d 3v (3.20)

so that

m n V V V V + (P V V )S + (Q V )S + R = m vvvvf (x, v,t)d 3v (3.21)

and
m n V V V V + (P V V V )S + (Q V V )S + (R V )S + S
(3.22)
= m vvvvvf
(x, v,t)d 3v

where the superscript s denotes a higher-rank symmetric tensor. An example of the

2nd-rank symmetric tensor is shown in (3.16).

Exercise 3.0.1
Let us define a generalized local entropy function S of the th species:
f
S = ln f d 3v + constant
n
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 29

Let f be a normal distribution function with number density n , temperature T ,

and zero average velocity. Determine how the entropy S varies with varying of the
number density n and the thermal pressure p .

In addition to the general definitions of the fluid variables, we shall encounter the following
integrations in deriving the fluid equations in the next section.

v {[E(x,t) + v B(x,t)] f (x,v,t)}d 3v = 0 (3.23)

v v {[E(x,t) + v B(x,t)] f (x,v,t)}d 3v = n (E + V B) (3.24)

vv v {[E(x,t) + v B(x,t)] f (x,v,t)}d 3v
(3.25)
= n (EV ) n [V (V B)]
S S
{(v V )[(v V ) B]}

S 3
f (x,v,t)d v

The following two integrations are useful for the higher moment integrations of the Vlasov
equation.

vvv v {[E(x,t) + v B(x,t)] f (x,v,t)}d 3v

n n
= n (EV V ) S (EP ) S n [V V (V B)]S [P (V B)]S (3.26)
m m
{(v V )(v V )[(v V ) B]}

S
f (x,v,t)d 3v

vvvv v {[E(x,t) + v B(x,t)] f (x,v,t)}d 3v

n n
= n (EV V V ) S (EP V ) S (EQ ) S
m m (3.27)
n n
n [V V V (V B)]S [P V (V B)]S [Q (V B)]S
m m
{(v V )(v V )(v V )[(v V ) B]}

S
f (x,v,t)d 3v

and so forth.
30 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Exercise 3.1.
Verify Eqs. (3.23), (3.24), (3.25), and (3.26).
Hint: There are three terms in the integration of Eq. (3.23), they are all in the following
form:

[ v E x f (x,v,t)dv x ]d 2v = E [ f x (x,v,t)]vv xx ==
+ 2
d v =0
x

There are nine terms in the integration of Eq. (3.24), in which six of them are in the
following form:

[v v
y E x f (x,v,t)dv x ]d 2v = v E [ f y x (x,v,t)]vv xx ==
+ 2
d v =0
x

and the rest three of them are in the following form:

[ v x
vx
E x f (x,v,t)dv x ]d 2v

= E {[v x f (x,v,t)]vv xx ==
x
+
f (x,v,t)dv x }d 2v = 0 E x n

There are 27 terms in the integration of Eq. (3.25), in which six of them are in the
following form:

[v v v
y z E x f (x,v,t)dv x ]d 2v = v v E [ fy z x (x,v,t)]vv xx ==
+ 2
d v =0
x

another six of them are in the following type:

[v v v
y y E x f (x,v,t)dv x ]d 2v = v v E [ fy y x (x,v,t)]vv xx ==
+ 2
d v =0
x

another twelve of them are in the following form:

[ v v y x
vx
E x f (x,v,t)dv x ]d 2v

= v E {[v y x x f (x,v,t)]vv xx ==
+
f (x,v,t)dv x }d 2v = 0 Vy E x n

and the rest three of them are in the following form:

[ v v x x
vx
E x f (x,v,t)dv x ]d 2v

= E {[v v x x x f (x,v,t)]vv xx ==
+
2v x f (x,v,t)dv x }d 2v
= 0 2E xVx n = (E xVx + Vx E x )n
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 31

3.3. The Fluid Equations

Fluid equations can be obtained from integration of the Vlasov equation in the velocity space.
For instance (e.g., Rossi and Olbert, 1970; Chao, 1970):

We can obtain the continuity equation of the th species from (3.1) d 3v .

We can obtain the momentum equation of the th species from m v(3.1) d 3v .

We can obtain the pressure equation of the th species from m vv(3.1) d 3v .

1
We can obtain the energy equation of the th species from 2 m v
2
(3.1) d 3v .

If we consider entire plasma system as a single fluid medium, the following integrations are
useful in obtaining one-fluid plasma equations.

m (3.1) d 3v yields the one-fluid mass continuity equation.

m v(3.1)
d 3v yields the one-fluid momentum equation.

m vv(3.1)
d 3v yields the one-fluid pressure equation.

1
2 m v (3.1)
2
d 3v yields the one-fluid energy equation.

e (3.1)
d 3v yields the one-fluid charge continuity equation.

e v(3.1)
d 3v yields the one-fluid Ohms law.

3.3.1. The Fluid Equations of the th Species

The continuity equation of the th species, obtained from (3.1 )d 3v , is

n
+ (n V ) = 0 (3.28)
t

The momentum equation of the th species, obtained from m v(3.1)d v , is

3

(m n V ) + (m n V V + P ) e n (E + V B) = 0 (3.29)
t

The pressure equation of the th species, obtained from m vv(3.1)d v , is

3

32 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

(m n V V + P ) + [m n V V V + (P V ) S + Q ]
t
e n (EV + V E) e n [V (V B) + (V B)V ] (3.30)
e {(v V )[(v V ) B] + [(v V ) B](v V )} f
(x,v,t)}d 3v = 0

1
The energy equation of the th species, obtained from 2 m v
2
(3.1)d 3v , is

1 3 1 3
( m nV2 + p ) + [( m nV2 + p )V + P V + q ] e n E V = 0 (3.31)
t 2 2 2 2
As we can see that the continuity equation is a scalar equation, the momentum equation
is a vector equation, the pressure equation is a second-rank-tensor equation, and the energy
equation is a scalar equation. Likewise, we can obtain a third-rank-tensor heat-flux

m vvv(3.1)d v
3
equation from , and a fourth-rank-tensor equation from

m vvvv(3.1)d v , and so forth.

3

If the th species has reached to a thermal-dynamic-equilibrium state, then we have

P = 1 p , and all the higher order rank tensors are vanished, i.e., Q = 0 , R = 0 ,
S = 0 , and so forth. In this case, the six-dimensional Vlasov equation can be replaced
by the three-dimensional equations (3.28), (3.29), and (3.31). Otherwise, we have to use
infinite number of three-dimensional equations to replace the six-dimensional Vlasov
equation. Thus, the fluid variables are meaningful and fluid equations are useful only when
the th species is in a thermal-dynamic-equilibrium state and can remain in a
quasi-thermal-dynamic-equilibrium state after they interaction with the waves.

3.3.2. The Two-Fluid Equations in the Convective-Time-Derivative Form

In fluid mechanics, it is commonly use dA /dt to denote the time derivatives of A(x,t)
along the trajectory of a fluid element in a velocity field V(x,t) . Namely,
dA /dt ( / t + V ) A , where V A is the convective time derivative of A(x,t) .
Equations obtained in the last section 3.3.1 can be rewritten in a convective-time-derivative
form. The convective-time-derivative term is a second-order small term in the linear wave
analysis of waves in a uniform background plasma. Thus, equations obtained in this section
are particularly useful in linear wave analysis.

The continuity equation (3.28) can be rewritten as

 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 33

( + V )n = n V (3.32)
t
The momentum equation (3.29) can be rewritten as

m n ( + V )V = P + e n (E + V B) (3.33)
t
The energy equation (3.31) can be rewritten as
3 3
[( + V ) p ] + p ( V ) + (P ) V + q = 0 (3.34)
2 t 2
For q = 0 and isotropic pressure P = 1 p , the energy equation (3.34) is reduced to the
the well-known adiabatic equation of state
3 d 3
[ ln( p n5 / 3 )] = [( + V )ln( p n5 / 3 )] = 0 (3.35)
2 dt 2 t

Exercise 3.2.
Verify Eqs. (3.33) and (3.34).

Answer to Exercise 3.2.

Proof of Eq. (3.33):
Substituting Eq. (3.28) into Eq. (3.29) yields

(m n V ) + (m n V V + P ) e n (E + V B)
t
n
= m V [ + (n V )] + m n ( + V )V + P e n (E + V B)
t t

= 0 + m n ( + V )V + P e n (E + V B) = 0
t
Proof of Eq. (3.34):
Substituting Eqs. (3.28) and (3.33) into Eq. (3.31), it yields
1 3 1 3
( m nV2 + p ) + [( m nV2 + p )V + P V + q ] e n E V
t 2 2 2 2
1 n
= mV2 [ + (n V )] + V [m n ( + V )V + P e n (E + V B)]
2 t t
3 3
+ [( + V ) p ] + p ( V ) + (P ) V + q
2 t 2
3 3
= 0 + 0 + [( + V ) p ] + p ( V ) + (P ) V + q = 0
2 t 2
34 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

In summary, Table 3.1 lists governing equations of ion-electron two-fluid plasma, in which
both species have reached to a thermal-dynamic-equilibrium state so that P = 1 p and
q = 0 . There are 16 unknowns ( n , V , p , E , B ) and 16 independent equations in this
system.
Table 3.1. The governing equations of the ion-electron two-fluid plasma
with an isotropic pressure and zero heat flux
SI Units Gaussian Units
The electrons equations: The electrons equations:

( + Ve )n e = n e Ve ( + Ve )n e = n e Ve
t t
V B
me n e ( + Ve )Ve = pe en e (E + Ve B) me n e ( + Ve )Ve = pe en e (E + e )
t t c

( + Ve )ln( pe n 5 /3
)=0 ( + Ve )ln( pe n 5 /3
)=0
t t
e e

The ions equations: The ions equations:

( + Vi )n i = n i Vi ( + Vi )n i = n i Vi
t t
V B
mi n i ( + Vi )Vi = pi + en i (E + Vi B) mi n i ( + Vi )Vi = pi + en i (E + i )
t t c

( + Vi )ln( pi n 5 /3
)=0 ( + Vi )ln( pi n 5 /3
)=0
t t
i i

The Maxwells equations: The Maxwells equations:

E = e(n i n e ) /0 E = 4 e(n i n e )
B=0 B=0
B 1 B
E = E =
t c t
1 E 4 1 E
B = 0e(n i Vi n e Ve ) + B = e(n i Vi n e Ve ) +
c2 t c c t
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 35

3.3.3. The One-Fluid Equations in the Conservative Form for Studying Nonlinear
Wave in the MHD Plasma

One-fluid equations in the conservative form (summarized in Table 3.2) will be obtained in
this section. Equations in conservative form are good for nonlinear wave analysis in the
magnetohydrodynamic (MHD) plasma. Before introducing the one-fluid plasma equations,
we shall first define one-fluid plasma variables.

3.3.3.1 The One-Fluid Variables

The mass density of the one-fluid plasma is defined by

(m n ) (3.36)

The charge density of the one-fluid plasma is defined by

c (e n ) (3.37)

The average (bulk) velocity of the one-fluid plasma is defined by

1
V (m n V )

(3.38)

The momentum of the one-fluid plasma is defined by

V (m n V ) (3.39)

The electric current density of the one-fluid plasma is defined by

J (e n V ) (3.40)

The total kinetic pressure of the one-fluid plasma is defined by

VV + P (m n V V + P ) (3.41)

where the thermal pressure tensor of the one-fluid plasma is defined by

P (m n V V + P ) VV (3.42)

or

P(x,t) [ m (v V)(v V) f (x, v,t)d 3v] (3.42')

The total kinetic pressure flux of the one-fluid plasma is defined by

36 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

VVV + (P V) S + Q [(m n V V V + (P V ) S + Q ] (3.43)

where the heat-flux tensor of the one-fluid plasma is defined by

Q [(m n V V V + (P V ) S + Q ] VVV (P V) S (3.44)

or

Q(x,t) [ m (v V)(v V)(v V) f (x, v,t)d 3v] (3.44')

The total kinetic energy flux of the one-fluid plasma is defined by

1 3 1 3
( V 2 + p)V + P V + q [( m nV2 + p )V + P V + q ] (3.45)
2 2 2 2
where the heat-flux vector of the one-fluid plasma is defined by
1 3 1 3
q [( m nV2 + p )V + P V + q ] ( V 2 + p)V P V (3.46)
2 2 2 2
or
1
q(x,t) [ m (v V) (v V)(v V) f (x,v,t)d 3v ] (3.46')
2

3.3.3.2 The One-Fluid Equations

The one-fluid mass continuity equation, obtained from m (3.1 ) d 3v , is

(m n ) + (m n V ) = 0
t

or

+ ( V) = 0 (3.47)
t

The one-fluid momentum equation, obtained from m v(3.1)

d 3v , is

(m n V ) + (m n V V + P ) [e n (E + V B)] = 0
t

or

( V) + ( VV + P) = c E + J B (3.48)
t
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 37

Substituting the Maxwells equations (3.2)~(3.5) into Eq. (3.48) yields

1 EB 1E 2 1B 2 BB
( V + 2 ) + ( VV + P + 0 0EE + )=0 (3.49)
t c 0 2 20 0
where (E B) / 0 is the Poynting vector of electromagnetic wave.

Exercise 3.3.
1E 2 BB 1B 2 1 EB
Show that c E + J B = (0EE 0 + ) 2 ( )
2 0 20 c t 0
Answer to Exercise 3.3.
c E + J B
1 E
= ( 0 E)E + [ ( B) 0 ] B
0 t
1 1 E
= ( 0 E)E + B B B 2 0 B
0 2 0 t
1 1 1
= ( 0 E)E + B B + B( B) B 2 0 (E B) + 0 (E B)
0 0 2 0 t t
BB 1B 2
= ( 0 E)E + ( ) 0 (E B) 0 E ( E)
0 2 0 t
BB 1B 2
= ( 0 E)E + ( ) 0 (E B) + 0 E E 0 E 2
0 2 0 t 2
1E 2 BB 1B 2 1 EB
= ( 0 EE 0 + ) 2 ( )
2 0 2 0 c t 0

1
The one-fluid energy equation, obtained from 2 m v
2
(3.1) d 3v , is

1 3 1 3
( m nV2 + p ) + [( m nV2 + p )V + P V + q ] (e n E V ) = 0
t 2 2 2 2

or
1 3 1 3
( V 2 + p) + [( V 2 + p)V + P V + q] = E J (3.50)
t 2 2 2 2

Substituting the Maxwells equations (3.2)~(3.5) into Eq. (3.50), it yields

1 3 E 2 B2 1 3 EB
( V 2 + p + 0 + ) + [( V 2 + p)V + P V + q + ]=0 (3.51)
t 2 2 2 20 2 2 0
38 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Exercise 3.4.
E B 0 E 2 B 2
Show that E J = ( ) ( + )
0 t 2 20

Answer to Exercise 3.4.

1 E 1 1 0 E 2
E J = E [ ( B) 0 ] = (E B) + B E
0 t 0 0 2 t
1 1 B 0 E 2 E B 0 E 2 B 2
= (E B) B = ( ) ( + )
0 0 t 2 t 0 t 2 20

The charge continuity, obtained from e (3.1)

d 3v , is

(e n ) + (e n V ) = 0
t

or

c + J = 0 (3.52)
t

The one-fluid Ohms law can be obtained from e v(3.1)

d 3v . There are two ways

to expand the integration e vv x f
d 3v , i.e.,

e vv x f
d 3v

= e [(v V ) + V ][(v V ) + V ] f d 3v

e
= ( P + e n V V )
m
JJ mi me mi me c
VJ + JV c VV [ + ]
e e en mi + me (mi + me ) 2 en
= { Pi Pe + }
mi me mi me c mi me c 2
1 ( )
mi + me en (mi + me ) 2 en
or
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 39

e vv x f
d 3v

= e [(v V) + V][(v V) + V] f d 3v

= [
e C.M.
P + (e Vn V e n VV) + (e n V V e n VV) + e n VV]
m
e C.M. e C.M.
= [ Pi Pe + VJ + JV c VV]
mi me

where n /(mi + me ) , and PC.M. m (v V)(v V) f (x, v,t)d 3v .

Thus, the one-fluid Ohms law can be written as

JJ mi me mi me c
VJ + JV c VV [ + ]
e e en mi + me (mi + me ) 2 en
J + { Pi Pe + }
t mi me mi me c mi me c 2
1 ( )
mi + me en (mi + me ) 2 en
e 2 e(mi me )
(E + V B) + ( c E + J B) = 0 (3.53)
mi me mi me
The one-fluid Ohms law can also be written as
e e C.M.
J + [ PiC.M. Pe + VJ + JV c VV]
t mi me
e 2 e(mi me )
(E + V B) + ( c E + J B) = 0 (3.54)
mi me mi me
Substituting the result of exercise 3.3 into equation (3.54), it yields
e(mi me ) 1 E B e e C.M.
[J ] + [ PiC.M. Pe + VJ + JV c VV
t m i m e c 0
2
mi me

e(mi me ) 1E 2 BB 1B 2 e 2
+ (0EE 0 + )] = (E + V B) (3.54')
mi me 2 0 20 mi me
Eq. (3.54) is commonly called the generalized Ohms law. Please see Appendix B for
discussion of the generalized Omhs Law in detail.

Exercise 3.5.
Verify Eqs. (3.53).
Hint: = mi ni + me ne , c = e(n i ne ) , V = min iVi + me ne Ve , and J = e(ni Vi ne Ve )
me,i m m
yields n i,e = ( + c ) /(mi + me ) , Vi,e = ( V + e.i J) /( + e,i c )
ei,e ei,e ei,e
40 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Table 3.2. The one-fluid equations in the conservative form

SI Units Gaussian Units
The mass continuity equation The mass continuity equation

+ ( V) = 0 + ( V) = 0
t t
The momentum equation The momentum equation
1 EB 1 EB
( V + 2 ) ( V + 2 )
t c 0 t c 4
1E 2 1B 2 BB 1E 2 EE 1B 2 BB
+ [ VV + P + 0 ( EE) + ]=0 + [ VV + P + + ]=0
2 20 0 8 4 8 4
The energy equation The energy equation
1 3 E B 2 2
1 3 E 2 B2
( V 2 + p + 0 + ) ( V 2 + p + + )
t 2 2 2 20 t 2 2 8 8
1 3 EB 1 3 EB
+ [( V 2 + p)V + P V + q + ]=0 + [( V 2 + p)V + P V + q + ]=0
2 2 0 2 2 4
The charge continuity equation The charge continuity equation

c + J = 0 c + J = 0
t t
The generalized Ohms law The generalized Ohms law
e(mi me ) 1 E B e(mi me ) 1 E B
[J ]+ [J ]+
t m i m e c 2 0 t mi me c 2 4
e e C.M. e e C.M.
[ PiC.M. Pe + VJ + JV c VV [ PiC.M. Pe + VJ + JV c VV
mi me mi me
e(mi me ) 1E 2 BB 1B 2 e(mi me ) EE 1E 2 BB 1B 2
+ (0EE 0 + )] + ( + )]
mi me 2 0 20 mi me 4 8 4 8
e 2 e 2 VB
= (E + V B) = (E + )
mi me mi me c
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 41

3.3.4. The One-Fluid Equations in the Convective-Time-Derivative Form

Equations obtained in the last section 3.3.3 can be rewritten in a convective-time-derivative

form. The convective-time-derivative term is a second-order small term in the linear wave
analysis of waves in a uniform background plasma. Thus, equations obtained in this section
are particularly useful in linear wave analysis.

The one-fluid continuity equation (3.47) can be rewritten as

( + V ) = V (3.55)
t
The one-fluid momentum equation (3.48) can be rewritten as

( + V )V = P + c E + J B (3.56)
t
The one-fluid energy equation (3.50) can be rewritten as
3 3
[( + V ) p] + p( V) + (P ) V + q + c E V J (E + V B) = 0 (3.57)
2 t 2

Exercise 3.6.
Verify Eqs. (3.56) and (3.57).

Answer to Exercise 3.6.

Proof of Eq. (3.56):
Substituting Eq. (3.47) into Eq. (3.48), it yields

( V) + ( VV + P) c E + J B
t

= V[ + ( V)] + ( + V )V + P c E J B
t t

= 0 + ( + V )V + P c E J B = 0
t
Proof of Eq. (3.57):
Substituting Eqs. (3.47) and (3.56) into Eq. (3.50), it yields
42 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

1 3 1 3
( V 2 + p) + [( V 2 + p)V + P V + q] E J
t 2 2 2 2
1
= V 2 [ + ( V)] + V [ ( + V )V + P c E J B]
2 t t
3 3
+ [( + V ) p] + p( V) + (P ) V + q + V ( c E + J B) E J
2 t 2
3 3
= 0 + 0 + [( + V ) p] + p( V) + (P ) V + q + c E V J (E + V B) = 0
2 t 2

Magnetohydrodynamic (MHD) phenomena are very low frequency and very long wavelength
phenomena in the plasma. The time scale of the MHD phenomena is equal or greater than
10 3 ions characteristic time scale, such as the ion gyro period ( 2 / ci ), or the ions plasma
oscillation period ( 2 / pi ). The spatial scale of the MHD phenomena is equal or greater
3
than 10 ions characteristic length, such as the ion gyro radius ( v 0 /ci ), or the ions inertial
length ( c / pi ).

The quasi-neutrality assumption

c ni ne
= 0
ne n
is applicable to the MHD plasma. Thus, we can assume c = 0 for the MHD plasma. As a
result, for MHD plasma, the charge continuity equation (3.52) can be reduced to
J =0 (3.58)
Since the curl of Eq. (3.5) is the charge continuity equation (3.52), the displacement current
in the Maxwells equation can be ignored if c / t = 0 . Thus, for the MHD plasma, Eq.
(3.5) is reduced to
B = 0 J (3.59)
In the low-frequency and long-wavelength limit, the generalized Ohms law is reduced to the
MHD Ohms law (see Appendix B),
E+ VB =0 (3.60)

The MHD Ohms law leads to the frozen-in flux in the MHD Plasma (see Appendix C).

For MHD plasma (which satisfies the MHD Ohms Law) with zero heat flux, the energy
equation (3.57) can be reduced to the adiabatic equation of state.
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 43

Case 1: The adiabatic equation of state of an isotropic MHD plasma

If P = 1 p , q = 0 , c = 0 , and E + V B = 0 , the energy equation (3.57) is reduced to
the following adiabatic equation of state
3 d 3
[ ln( p 5 / 3 )] = [( + V )ln( p 5 / 3 )] = 0 (3.61)
2 dt 2 t

Case 2: The adiabatic condition of an anisotropic MHD plasma

If P = e ||e || p|| + (1 e||e || )p = 1 p + e||e || (p|| p ) , q = 0 , c = 0 , and E + V B = 0 ,
where e || = B/ B , the energy equation (3.57) is reduced to the following form
p|| d p|| B 2 d p
[ ln( 3 )] + p [ ln( )] = 0 (3.62)
2 dt dt B
or
p|| p B2 p
[( + V )ln( || 3 )] + p [( + V )ln( )] = 0 (3.62')
2 t t B

Under an assumption that no momentum exchange in the directions parallel to and

perpendicular to the local magnetic field, Chew et al. (1956) obtained the well-known
Chew-Goldberger-Low theory or the CGL double adiabatic equation of states, i.e.,
d( p|| B 2 / 3 )
=0 (3.63)
dt
and
d( p / B)
=0 (3.64)
dt

Eqs. (3.63) and (3.64) are a special set of solutions of equation (3.62). It can be shown that
Chew-Goldberger-Low theory is only applicable to a system with uniform magnetic field
strength, so that the mirror motion is prohibited in the system. Professor J. K. Chao has
found an example from magnetosheath observations along the Sun-Earth line, which shows
that Eq. (3.62) is a more general adiabatic condition than the Chew-Goldberger-Low theory.
44 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Figure 3.1. Magnetosheath observations along the Sun-Earth line. Panel (a) is a plot of
p [d ln( p / B)2 / dt] . Panel (b) is a plot of ( p|| )[d ln( p|| B 2 / 3 ) / dt] . Panel (c) is a plot

of p [d ln( p / B)2 / dt] + p|| [d ln( p|| B 2 / 3 ) / dt] . Wave amplitude in Panel (c) is much
smaller than the wave amplitude in Panels (a) and (b). Results shown in this figure indicate
that Eq. (3.62) is a more general adiabatic condition than the Chew-Goldberger-Low theory
or the so-called CGL double adiabatic equation of states (Chew et al., 1956). Observational
data shown in this figure are obtained from IRM/AMPTE at 0645UT-0715UT on September
1, 1984. (Courtesy of Professor J. K. Chao)

Figure 3.1 shows the magnetosheath observations along the Sun-Earth line. Panel (a) is a plot
of p [d ln( p / B)2 / dt] . Panel (b) is a plot of ( p|| )[d ln( p|| B 2 / 3 ) / dt] . Panel (c) is a

plot of p [d ln( p / B)2 / dt] + p|| [d ln( p|| B 2 / 3 ) / dt] . Wave amplitude in Panel (c) is
much smaller than the wave amplitude in Panels (a) and (b). These results indicate that Eq.
(3.62) is a more general adiabatic condition than the CGL double adiabatic equation of states
(Chew et al., 1956).

In summary, Table 3.3 lists the governing equations of the MHD plasma with an isotropic
pressure ( P = 1 p ) and zero heat flux ( q = 0 ). There are 14 unknowns ( n , V , p , J , E , B ) and
14 independent equations in this system.
 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation 45

Table 3.3. Governing equations of MHD plasma with isotropic pressure and zero heat flux
SI Units Gaussian Units
The mass continuity equation The mass continuity equation

( + V ) = V ( + V ) = V
t t
The MHD momentum equation The MHD momentum equation
JB
( + V )V = p + J B ( + V )V = p +
t t c
The MHD energy equation The MHD energy equation
3 3
[( + V )ln( p 5 / 3 )] = 0 [( + V )ln( p 5 / 3 )] = 0
2 t 2 t
The MHD charge continuity equation The MHD charge continuity equation
J=0 J =0
The MHD Ohms law
The MHD Ohms law
VB
E+ VB =0 E+ =0
c
The Maxwells equations: The Maxwells equations:
E 0 E 0
B=0 B=0
B 1 B
E = E =
t c t
B = 0 J B = (4 /c)J

References

Chao, J. K. (1970), Interplanetary Collisionless Shock Waves, Rep. CSR TR-70-3, Mass. Inst.
of Technol. Cent. for Space Res., Cambridge, Mass..
Chen, F. F. (1984), Introduction to Plasma Physics and Controlled Fusion, Volume 1:
Plasma Physics, 2nd edition, Plenum Press, New York.
Chew, C. F., M. L. Goldberger, and F. E. Low (1956), The Boltzmann equation and the
one-fluid hydromagnetic equations in the absence of particle collisions, Proc. Roy. Soc.
London Ser. A, 236, 112.
46 Chapter 3. Deriving the Fluid Equations From the Vlasov Equation

Krall, N. A., and A. W. Trivelpiece (1973), Principles of Plasma Physics, McGraw-Hill

Book Company, New York.
Nicholson, D. R. (1983), Introduction to Plasma Theory, John Wiley & Sons, New York.
Rossi, B., and S. Olbert (1970), Introduction to the Physics of Space, McGraw-Hill, New
York.